Upload folder using huggingface_hub
Browse files- checkpoint-7000/config.json +67 -0
- checkpoint-7000/model.safetensors +3 -0
- checkpoint-7000/optimizer.pt +3 -0
- checkpoint-7000/rng_state.pth +3 -0
- checkpoint-7000/scaler.pt +3 -0
- checkpoint-7000/scheduler.pt +3 -0
- checkpoint-7000/trainer_state.json +1140 -0
- checkpoint-7000/training_args.bin +3 -0
- checkpoint-9048/config.json +67 -0
- checkpoint-9048/model.safetensors +3 -0
- checkpoint-9048/optimizer.pt +3 -0
- checkpoint-9048/rng_state.pth +3 -0
- checkpoint-9048/scaler.pt +3 -0
- checkpoint-9048/scheduler.pt +3 -0
- checkpoint-9048/trainer_state.json +1456 -0
- checkpoint-9048/training_args.bin +3 -0
- config.json +67 -0
- model.safetensors +3 -0
- preprocessor_config.json +23 -0
- training_args.bin +3 -0
checkpoint-7000/config.json
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_num_labels": 2,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Swinv2ForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"depths": [
|
| 8 |
+
2,
|
| 9 |
+
2,
|
| 10 |
+
18,
|
| 11 |
+
2
|
| 12 |
+
],
|
| 13 |
+
"drop_path_rate": 0.1,
|
| 14 |
+
"dtype": "float32",
|
| 15 |
+
"embed_dim": 192,
|
| 16 |
+
"encoder_stride": 32,
|
| 17 |
+
"hidden_act": "gelu",
|
| 18 |
+
"hidden_dropout_prob": 0.0,
|
| 19 |
+
"hidden_size": 1536,
|
| 20 |
+
"id2label": {
|
| 21 |
+
"0": "artificial",
|
| 22 |
+
"1": "real"
|
| 23 |
+
},
|
| 24 |
+
"image_size": 256,
|
| 25 |
+
"initializer_range": 0.02,
|
| 26 |
+
"label2id": {
|
| 27 |
+
"artificial": 0,
|
| 28 |
+
"real": 1
|
| 29 |
+
},
|
| 30 |
+
"layer_norm_eps": 1e-05,
|
| 31 |
+
"mlp_ratio": 4.0,
|
| 32 |
+
"model_type": "swinv2",
|
| 33 |
+
"num_channels": 3,
|
| 34 |
+
"num_heads": [
|
| 35 |
+
6,
|
| 36 |
+
12,
|
| 37 |
+
24,
|
| 38 |
+
48
|
| 39 |
+
],
|
| 40 |
+
"num_layers": 4,
|
| 41 |
+
"out_features": [
|
| 42 |
+
"stage4"
|
| 43 |
+
],
|
| 44 |
+
"out_indices": [
|
| 45 |
+
4
|
| 46 |
+
],
|
| 47 |
+
"patch_size": 4,
|
| 48 |
+
"path_norm": true,
|
| 49 |
+
"pretrained_window_sizes": [
|
| 50 |
+
12,
|
| 51 |
+
12,
|
| 52 |
+
12,
|
| 53 |
+
6
|
| 54 |
+
],
|
| 55 |
+
"problem_type": "single_label_classification",
|
| 56 |
+
"qkv_bias": true,
|
| 57 |
+
"stage_names": [
|
| 58 |
+
"stem",
|
| 59 |
+
"stage1",
|
| 60 |
+
"stage2",
|
| 61 |
+
"stage3",
|
| 62 |
+
"stage4"
|
| 63 |
+
],
|
| 64 |
+
"transformers_version": "4.57.6",
|
| 65 |
+
"use_absolute_embeddings": false,
|
| 66 |
+
"window_size": 16
|
| 67 |
+
}
|
checkpoint-7000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea8282e33616dd8c5aa9f5867ae40e89e0fe0aa87ae98770e378aa3b88642bdd
|
| 3 |
+
size 780886280
|
checkpoint-7000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b92ea49225fdac9e53340255cfc0bcbc04dd97cd22c8378cb41fafe01c9db2de
|
| 3 |
+
size 1562050030
|
checkpoint-7000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:25216cdbdd6d5b73497550f154dc6f2a31108bdd36ee6ea1120cf1874649a729
|
| 3 |
+
size 14645
|
checkpoint-7000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d134f260f8b6b79c56f3e4f54c5eeeb4fe7f1af8195ea5159f9f83dcb8c6e23b
|
| 3 |
+
size 1383
|
checkpoint-7000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b61ebc5917f935598f7357b21c31de550228283ee833d23f11059ae1a12ea0eb
|
| 3 |
+
size 1465
|
checkpoint-7000/trainer_state.json
ADDED
|
@@ -0,0 +1,1140 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 7000,
|
| 3 |
+
"best_metric": 0.9755905511811024,
|
| 4 |
+
"best_model_checkpoint": "output/checkpoint-7000",
|
| 5 |
+
"epoch": 2.320954907161804,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 7000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.016578249336870028,
|
| 14 |
+
"grad_norm": 15.383491516113281,
|
| 15 |
+
"learning_rate": 2.1633554083885212e-06,
|
| 16 |
+
"loss": 1.0204,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.033156498673740056,
|
| 21 |
+
"grad_norm": 15.810857772827148,
|
| 22 |
+
"learning_rate": 4.370860927152319e-06,
|
| 23 |
+
"loss": 0.8263,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.04973474801061008,
|
| 28 |
+
"grad_norm": 29.156631469726562,
|
| 29 |
+
"learning_rate": 6.578366445916116e-06,
|
| 30 |
+
"loss": 0.7179,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.06631299734748011,
|
| 35 |
+
"grad_norm": 13.032061576843262,
|
| 36 |
+
"learning_rate": 8.785871964679912e-06,
|
| 37 |
+
"loss": 0.4882,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.08289124668435013,
|
| 42 |
+
"grad_norm": 7.32506799697876,
|
| 43 |
+
"learning_rate": 1.099337748344371e-05,
|
| 44 |
+
"loss": 0.4156,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.09946949602122016,
|
| 49 |
+
"grad_norm": 16.674861907958984,
|
| 50 |
+
"learning_rate": 1.3200883002207508e-05,
|
| 51 |
+
"loss": 0.3548,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.11604774535809019,
|
| 56 |
+
"grad_norm": 39.721839904785156,
|
| 57 |
+
"learning_rate": 1.5408388520971304e-05,
|
| 58 |
+
"loss": 0.3145,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.13262599469496023,
|
| 63 |
+
"grad_norm": 9.635961532592773,
|
| 64 |
+
"learning_rate": 1.76158940397351e-05,
|
| 65 |
+
"loss": 0.292,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.14920424403183025,
|
| 70 |
+
"grad_norm": 3.8946573734283447,
|
| 71 |
+
"learning_rate": 1.9823399558498897e-05,
|
| 72 |
+
"loss": 0.3039,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.16578249336870027,
|
| 77 |
+
"grad_norm": 2.919046401977539,
|
| 78 |
+
"learning_rate": 1.989296102385108e-05,
|
| 79 |
+
"loss": 0.2764,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16578249336870027,
|
| 84 |
+
"eval_accuracy": 0.9015748031496063,
|
| 85 |
+
"eval_loss": 0.25162026286125183,
|
| 86 |
+
"eval_runtime": 8.6404,
|
| 87 |
+
"eval_samples_per_second": 293.967,
|
| 88 |
+
"eval_steps_per_second": 9.259,
|
| 89 |
+
"step": 500
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.1823607427055703,
|
| 93 |
+
"grad_norm": 4.6071295738220215,
|
| 94 |
+
"learning_rate": 1.9776614310645726e-05,
|
| 95 |
+
"loss": 0.2176,
|
| 96 |
+
"step": 550
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.1989389920424403,
|
| 100 |
+
"grad_norm": 1.5469706058502197,
|
| 101 |
+
"learning_rate": 1.9660267597440373e-05,
|
| 102 |
+
"loss": 0.2785,
|
| 103 |
+
"step": 600
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.21551724137931033,
|
| 107 |
+
"grad_norm": 18.289094924926758,
|
| 108 |
+
"learning_rate": 1.9543920884235023e-05,
|
| 109 |
+
"loss": 0.1978,
|
| 110 |
+
"step": 650
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.23209549071618038,
|
| 114 |
+
"grad_norm": 16.161046981811523,
|
| 115 |
+
"learning_rate": 1.942757417102967e-05,
|
| 116 |
+
"loss": 0.2398,
|
| 117 |
+
"step": 700
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.2486737400530504,
|
| 121 |
+
"grad_norm": 4.511693000793457,
|
| 122 |
+
"learning_rate": 1.9311227457824316e-05,
|
| 123 |
+
"loss": 0.1964,
|
| 124 |
+
"step": 750
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.26525198938992045,
|
| 128 |
+
"grad_norm": 0.6140362620353699,
|
| 129 |
+
"learning_rate": 1.9194880744618966e-05,
|
| 130 |
+
"loss": 0.1817,
|
| 131 |
+
"step": 800
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.28183023872679047,
|
| 135 |
+
"grad_norm": 34.74034881591797,
|
| 136 |
+
"learning_rate": 1.9078534031413613e-05,
|
| 137 |
+
"loss": 0.1988,
|
| 138 |
+
"step": 850
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.2984084880636605,
|
| 142 |
+
"grad_norm": 2.9800262451171875,
|
| 143 |
+
"learning_rate": 1.8962187318208263e-05,
|
| 144 |
+
"loss": 0.1641,
|
| 145 |
+
"step": 900
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.3149867374005305,
|
| 149 |
+
"grad_norm": 15.570615768432617,
|
| 150 |
+
"learning_rate": 1.884584060500291e-05,
|
| 151 |
+
"loss": 0.1554,
|
| 152 |
+
"step": 950
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.33156498673740054,
|
| 156 |
+
"grad_norm": 1.9845480918884277,
|
| 157 |
+
"learning_rate": 1.872949389179756e-05,
|
| 158 |
+
"loss": 0.1855,
|
| 159 |
+
"step": 1000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.33156498673740054,
|
| 163 |
+
"eval_accuracy": 0.9358267716535433,
|
| 164 |
+
"eval_loss": 0.18621081113815308,
|
| 165 |
+
"eval_runtime": 8.3497,
|
| 166 |
+
"eval_samples_per_second": 304.203,
|
| 167 |
+
"eval_steps_per_second": 9.581,
|
| 168 |
+
"step": 1000
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.34814323607427056,
|
| 172 |
+
"grad_norm": 14.85474681854248,
|
| 173 |
+
"learning_rate": 1.8613147178592206e-05,
|
| 174 |
+
"loss": 0.1725,
|
| 175 |
+
"step": 1050
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.3647214854111406,
|
| 179 |
+
"grad_norm": 4.149899959564209,
|
| 180 |
+
"learning_rate": 1.8496800465386856e-05,
|
| 181 |
+
"loss": 0.1513,
|
| 182 |
+
"step": 1100
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.3812997347480106,
|
| 186 |
+
"grad_norm": 22.695873260498047,
|
| 187 |
+
"learning_rate": 1.8380453752181503e-05,
|
| 188 |
+
"loss": 0.1767,
|
| 189 |
+
"step": 1150
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.3978779840848806,
|
| 193 |
+
"grad_norm": 0.17689716815948486,
|
| 194 |
+
"learning_rate": 1.8264107038976153e-05,
|
| 195 |
+
"loss": 0.1623,
|
| 196 |
+
"step": 1200
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.41445623342175064,
|
| 200 |
+
"grad_norm": 30.613101959228516,
|
| 201 |
+
"learning_rate": 1.81477603257708e-05,
|
| 202 |
+
"loss": 0.1682,
|
| 203 |
+
"step": 1250
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.43103448275862066,
|
| 207 |
+
"grad_norm": 16.982980728149414,
|
| 208 |
+
"learning_rate": 1.8031413612565446e-05,
|
| 209 |
+
"loss": 0.1508,
|
| 210 |
+
"step": 1300
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.44761273209549074,
|
| 214 |
+
"grad_norm": 0.1425020545721054,
|
| 215 |
+
"learning_rate": 1.7915066899360092e-05,
|
| 216 |
+
"loss": 0.1679,
|
| 217 |
+
"step": 1350
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.46419098143236076,
|
| 221 |
+
"grad_norm": 5.924993991851807,
|
| 222 |
+
"learning_rate": 1.7798720186154742e-05,
|
| 223 |
+
"loss": 0.1387,
|
| 224 |
+
"step": 1400
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.4807692307692308,
|
| 228 |
+
"grad_norm": 5.42221212387085,
|
| 229 |
+
"learning_rate": 1.768237347294939e-05,
|
| 230 |
+
"loss": 0.1627,
|
| 231 |
+
"step": 1450
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.4973474801061008,
|
| 235 |
+
"grad_norm": 31.555957794189453,
|
| 236 |
+
"learning_rate": 1.756602675974404e-05,
|
| 237 |
+
"loss": 0.1188,
|
| 238 |
+
"step": 1500
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.4973474801061008,
|
| 242 |
+
"eval_accuracy": 0.9417322834645669,
|
| 243 |
+
"eval_loss": 0.1927609145641327,
|
| 244 |
+
"eval_runtime": 8.4174,
|
| 245 |
+
"eval_samples_per_second": 301.756,
|
| 246 |
+
"eval_steps_per_second": 9.504,
|
| 247 |
+
"step": 1500
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5139257294429708,
|
| 251 |
+
"grad_norm": 15.842325210571289,
|
| 252 |
+
"learning_rate": 1.7449680046538686e-05,
|
| 253 |
+
"loss": 0.1678,
|
| 254 |
+
"step": 1550
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5305039787798409,
|
| 258 |
+
"grad_norm": 6.040674686431885,
|
| 259 |
+
"learning_rate": 1.7333333333333336e-05,
|
| 260 |
+
"loss": 0.1215,
|
| 261 |
+
"step": 1600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.5470822281167109,
|
| 265 |
+
"grad_norm": 0.5153898596763611,
|
| 266 |
+
"learning_rate": 1.7216986620127982e-05,
|
| 267 |
+
"loss": 0.1507,
|
| 268 |
+
"step": 1650
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.5636604774535809,
|
| 272 |
+
"grad_norm": 18.98447608947754,
|
| 273 |
+
"learning_rate": 1.7100639906922632e-05,
|
| 274 |
+
"loss": 0.1481,
|
| 275 |
+
"step": 1700
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.580238726790451,
|
| 279 |
+
"grad_norm": 43.290855407714844,
|
| 280 |
+
"learning_rate": 1.698429319371728e-05,
|
| 281 |
+
"loss": 0.1422,
|
| 282 |
+
"step": 1750
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.596816976127321,
|
| 286 |
+
"grad_norm": 0.14761729538440704,
|
| 287 |
+
"learning_rate": 1.686794648051193e-05,
|
| 288 |
+
"loss": 0.1551,
|
| 289 |
+
"step": 1800
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.613395225464191,
|
| 293 |
+
"grad_norm": 96.18028259277344,
|
| 294 |
+
"learning_rate": 1.6751599767306575e-05,
|
| 295 |
+
"loss": 0.1303,
|
| 296 |
+
"step": 1850
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.629973474801061,
|
| 300 |
+
"grad_norm": 0.17378291487693787,
|
| 301 |
+
"learning_rate": 1.6635253054101222e-05,
|
| 302 |
+
"loss": 0.1484,
|
| 303 |
+
"step": 1900
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.646551724137931,
|
| 307 |
+
"grad_norm": 0.2519828677177429,
|
| 308 |
+
"learning_rate": 1.6518906340895872e-05,
|
| 309 |
+
"loss": 0.1845,
|
| 310 |
+
"step": 1950
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.6631299734748011,
|
| 314 |
+
"grad_norm": 37.454341888427734,
|
| 315 |
+
"learning_rate": 1.640255962769052e-05,
|
| 316 |
+
"loss": 0.1544,
|
| 317 |
+
"step": 2000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.6631299734748011,
|
| 321 |
+
"eval_accuracy": 0.9551181102362205,
|
| 322 |
+
"eval_loss": 0.14238852262496948,
|
| 323 |
+
"eval_runtime": 8.4401,
|
| 324 |
+
"eval_samples_per_second": 300.943,
|
| 325 |
+
"eval_steps_per_second": 9.479,
|
| 326 |
+
"step": 2000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.6797082228116711,
|
| 330 |
+
"grad_norm": 9.755090713500977,
|
| 331 |
+
"learning_rate": 1.6286212914485165e-05,
|
| 332 |
+
"loss": 0.1176,
|
| 333 |
+
"step": 2050
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.6962864721485411,
|
| 337 |
+
"grad_norm": 0.9808468222618103,
|
| 338 |
+
"learning_rate": 1.6169866201279815e-05,
|
| 339 |
+
"loss": 0.1234,
|
| 340 |
+
"step": 2100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.7128647214854111,
|
| 344 |
+
"grad_norm": 4.4084649085998535,
|
| 345 |
+
"learning_rate": 1.6053519488074462e-05,
|
| 346 |
+
"loss": 0.0882,
|
| 347 |
+
"step": 2150
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.7294429708222812,
|
| 351 |
+
"grad_norm": 35.67166519165039,
|
| 352 |
+
"learning_rate": 1.5937172774869112e-05,
|
| 353 |
+
"loss": 0.1246,
|
| 354 |
+
"step": 2200
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.7460212201591512,
|
| 358 |
+
"grad_norm": 54.88447189331055,
|
| 359 |
+
"learning_rate": 1.582082606166376e-05,
|
| 360 |
+
"loss": 0.1504,
|
| 361 |
+
"step": 2250
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.7625994694960212,
|
| 365 |
+
"grad_norm": 0.32783272862434387,
|
| 366 |
+
"learning_rate": 1.570447934845841e-05,
|
| 367 |
+
"loss": 0.1419,
|
| 368 |
+
"step": 2300
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.7791777188328912,
|
| 372 |
+
"grad_norm": 0.9903936386108398,
|
| 373 |
+
"learning_rate": 1.5588132635253055e-05,
|
| 374 |
+
"loss": 0.1154,
|
| 375 |
+
"step": 2350
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.7957559681697612,
|
| 379 |
+
"grad_norm": 0.5903331637382507,
|
| 380 |
+
"learning_rate": 1.5471785922047705e-05,
|
| 381 |
+
"loss": 0.0817,
|
| 382 |
+
"step": 2400
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.8123342175066313,
|
| 386 |
+
"grad_norm": 6.974926471710205,
|
| 387 |
+
"learning_rate": 1.5355439208842352e-05,
|
| 388 |
+
"loss": 0.1164,
|
| 389 |
+
"step": 2450
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.8289124668435013,
|
| 393 |
+
"grad_norm": 25.492650985717773,
|
| 394 |
+
"learning_rate": 1.5239092495637e-05,
|
| 395 |
+
"loss": 0.1592,
|
| 396 |
+
"step": 2500
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.8289124668435013,
|
| 400 |
+
"eval_accuracy": 0.9641732283464567,
|
| 401 |
+
"eval_loss": 0.131427600979805,
|
| 402 |
+
"eval_runtime": 8.5238,
|
| 403 |
+
"eval_samples_per_second": 297.989,
|
| 404 |
+
"eval_steps_per_second": 9.385,
|
| 405 |
+
"step": 2500
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.8454907161803713,
|
| 409 |
+
"grad_norm": 24.847103118896484,
|
| 410 |
+
"learning_rate": 1.5122745782431647e-05,
|
| 411 |
+
"loss": 0.1119,
|
| 412 |
+
"step": 2550
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.8620689655172413,
|
| 416 |
+
"grad_norm": 2.9396579265594482,
|
| 417 |
+
"learning_rate": 1.5006399069226297e-05,
|
| 418 |
+
"loss": 0.0978,
|
| 419 |
+
"step": 2600
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.8786472148541115,
|
| 423 |
+
"grad_norm": 21.483346939086914,
|
| 424 |
+
"learning_rate": 1.4890052356020943e-05,
|
| 425 |
+
"loss": 0.1233,
|
| 426 |
+
"step": 2650
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.8952254641909815,
|
| 430 |
+
"grad_norm": 0.2358592301607132,
|
| 431 |
+
"learning_rate": 1.4773705642815592e-05,
|
| 432 |
+
"loss": 0.1582,
|
| 433 |
+
"step": 2700
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.9118037135278515,
|
| 437 |
+
"grad_norm": 0.7912116646766663,
|
| 438 |
+
"learning_rate": 1.465735892961024e-05,
|
| 439 |
+
"loss": 0.1432,
|
| 440 |
+
"step": 2750
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.9283819628647215,
|
| 444 |
+
"grad_norm": 0.2763047516345978,
|
| 445 |
+
"learning_rate": 1.4541012216404888e-05,
|
| 446 |
+
"loss": 0.0988,
|
| 447 |
+
"step": 2800
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.9449602122015915,
|
| 451 |
+
"grad_norm": 0.904279887676239,
|
| 452 |
+
"learning_rate": 1.4424665503199535e-05,
|
| 453 |
+
"loss": 0.1295,
|
| 454 |
+
"step": 2850
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 0.9615384615384616,
|
| 458 |
+
"grad_norm": 0.578948974609375,
|
| 459 |
+
"learning_rate": 1.4308318789994185e-05,
|
| 460 |
+
"loss": 0.0949,
|
| 461 |
+
"step": 2900
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.9781167108753316,
|
| 465 |
+
"grad_norm": 5.489175796508789,
|
| 466 |
+
"learning_rate": 1.4191972076788831e-05,
|
| 467 |
+
"loss": 0.1224,
|
| 468 |
+
"step": 2950
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 0.9946949602122016,
|
| 472 |
+
"grad_norm": 17.271835327148438,
|
| 473 |
+
"learning_rate": 1.4075625363583481e-05,
|
| 474 |
+
"loss": 0.1476,
|
| 475 |
+
"step": 3000
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.9946949602122016,
|
| 479 |
+
"eval_accuracy": 0.9535433070866142,
|
| 480 |
+
"eval_loss": 0.16435901820659637,
|
| 481 |
+
"eval_runtime": 8.2931,
|
| 482 |
+
"eval_samples_per_second": 306.28,
|
| 483 |
+
"eval_steps_per_second": 9.647,
|
| 484 |
+
"step": 3000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 1.0112732095490715,
|
| 488 |
+
"grad_norm": 17.650592803955078,
|
| 489 |
+
"learning_rate": 1.3959278650378128e-05,
|
| 490 |
+
"loss": 0.064,
|
| 491 |
+
"step": 3050
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 1.0278514588859415,
|
| 495 |
+
"grad_norm": 0.08039892464876175,
|
| 496 |
+
"learning_rate": 1.3842931937172776e-05,
|
| 497 |
+
"loss": 0.055,
|
| 498 |
+
"step": 3100
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 1.0444297082228116,
|
| 502 |
+
"grad_norm": 36.28068542480469,
|
| 503 |
+
"learning_rate": 1.3726585223967423e-05,
|
| 504 |
+
"loss": 0.052,
|
| 505 |
+
"step": 3150
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.0610079575596818,
|
| 509 |
+
"grad_norm": 0.07026328891515732,
|
| 510 |
+
"learning_rate": 1.3610238510762073e-05,
|
| 511 |
+
"loss": 0.0706,
|
| 512 |
+
"step": 3200
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.0775862068965518,
|
| 516 |
+
"grad_norm": 6.0042853355407715,
|
| 517 |
+
"learning_rate": 1.349389179755672e-05,
|
| 518 |
+
"loss": 0.1037,
|
| 519 |
+
"step": 3250
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.0941644562334218,
|
| 523 |
+
"grad_norm": 0.8829227089881897,
|
| 524 |
+
"learning_rate": 1.337754508435137e-05,
|
| 525 |
+
"loss": 0.0536,
|
| 526 |
+
"step": 3300
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.1107427055702919,
|
| 530 |
+
"grad_norm": 54.231544494628906,
|
| 531 |
+
"learning_rate": 1.3261198371146016e-05,
|
| 532 |
+
"loss": 0.0374,
|
| 533 |
+
"step": 3350
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 1.1273209549071619,
|
| 537 |
+
"grad_norm": 0.07569246739149094,
|
| 538 |
+
"learning_rate": 1.3144851657940664e-05,
|
| 539 |
+
"loss": 0.0738,
|
| 540 |
+
"step": 3400
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 1.143899204244032,
|
| 544 |
+
"grad_norm": 0.017937984317541122,
|
| 545 |
+
"learning_rate": 1.3028504944735311e-05,
|
| 546 |
+
"loss": 0.03,
|
| 547 |
+
"step": 3450
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 1.160477453580902,
|
| 551 |
+
"grad_norm": 9.184370040893555,
|
| 552 |
+
"learning_rate": 1.2912158231529961e-05,
|
| 553 |
+
"loss": 0.0778,
|
| 554 |
+
"step": 3500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.160477453580902,
|
| 558 |
+
"eval_accuracy": 0.9677165354330709,
|
| 559 |
+
"eval_loss": 0.1568383127450943,
|
| 560 |
+
"eval_runtime": 8.2686,
|
| 561 |
+
"eval_samples_per_second": 307.187,
|
| 562 |
+
"eval_steps_per_second": 9.675,
|
| 563 |
+
"step": 3500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.177055702917772,
|
| 567 |
+
"grad_norm": 31.2453556060791,
|
| 568 |
+
"learning_rate": 1.2795811518324608e-05,
|
| 569 |
+
"loss": 0.04,
|
| 570 |
+
"step": 3550
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.193633952254642,
|
| 574 |
+
"grad_norm": 0.031678102910518646,
|
| 575 |
+
"learning_rate": 1.2679464805119258e-05,
|
| 576 |
+
"loss": 0.0687,
|
| 577 |
+
"step": 3600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.210212201591512,
|
| 581 |
+
"grad_norm": 0.20173047482967377,
|
| 582 |
+
"learning_rate": 1.2563118091913904e-05,
|
| 583 |
+
"loss": 0.092,
|
| 584 |
+
"step": 3650
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.226790450928382,
|
| 588 |
+
"grad_norm": 0.08714820444583893,
|
| 589 |
+
"learning_rate": 1.2446771378708553e-05,
|
| 590 |
+
"loss": 0.0692,
|
| 591 |
+
"step": 3700
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.243368700265252,
|
| 595 |
+
"grad_norm": 22.537704467773438,
|
| 596 |
+
"learning_rate": 1.23304246655032e-05,
|
| 597 |
+
"loss": 0.0652,
|
| 598 |
+
"step": 3750
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.259946949602122,
|
| 602 |
+
"grad_norm": 8.494444847106934,
|
| 603 |
+
"learning_rate": 1.2214077952297849e-05,
|
| 604 |
+
"loss": 0.0541,
|
| 605 |
+
"step": 3800
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.276525198938992,
|
| 609 |
+
"grad_norm": 0.4346514046192169,
|
| 610 |
+
"learning_rate": 1.2097731239092496e-05,
|
| 611 |
+
"loss": 0.055,
|
| 612 |
+
"step": 3850
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.293103448275862,
|
| 616 |
+
"grad_norm": 0.08467918634414673,
|
| 617 |
+
"learning_rate": 1.1981384525887146e-05,
|
| 618 |
+
"loss": 0.0454,
|
| 619 |
+
"step": 3900
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.3096816976127321,
|
| 623 |
+
"grad_norm": 0.10869591683149338,
|
| 624 |
+
"learning_rate": 1.1865037812681792e-05,
|
| 625 |
+
"loss": 0.0438,
|
| 626 |
+
"step": 3950
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.3262599469496021,
|
| 630 |
+
"grad_norm": 20.05512046813965,
|
| 631 |
+
"learning_rate": 1.174869109947644e-05,
|
| 632 |
+
"loss": 0.0661,
|
| 633 |
+
"step": 4000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.3262599469496021,
|
| 637 |
+
"eval_accuracy": 0.9696850393700788,
|
| 638 |
+
"eval_loss": 0.14853054285049438,
|
| 639 |
+
"eval_runtime": 8.4299,
|
| 640 |
+
"eval_samples_per_second": 301.307,
|
| 641 |
+
"eval_steps_per_second": 9.49,
|
| 642 |
+
"step": 4000
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 1.3428381962864722,
|
| 646 |
+
"grad_norm": 0.03564869612455368,
|
| 647 |
+
"learning_rate": 1.1632344386271089e-05,
|
| 648 |
+
"loss": 0.0529,
|
| 649 |
+
"step": 4050
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 1.3594164456233422,
|
| 653 |
+
"grad_norm": 0.020624179393053055,
|
| 654 |
+
"learning_rate": 1.1515997673065737e-05,
|
| 655 |
+
"loss": 0.0525,
|
| 656 |
+
"step": 4100
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.3759946949602122,
|
| 660 |
+
"grad_norm": 10.6432466506958,
|
| 661 |
+
"learning_rate": 1.1399650959860384e-05,
|
| 662 |
+
"loss": 0.0502,
|
| 663 |
+
"step": 4150
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 1.3925729442970822,
|
| 667 |
+
"grad_norm": 0.02102123387157917,
|
| 668 |
+
"learning_rate": 1.1283304246655034e-05,
|
| 669 |
+
"loss": 0.0582,
|
| 670 |
+
"step": 4200
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.4091511936339522,
|
| 674 |
+
"grad_norm": 32.820335388183594,
|
| 675 |
+
"learning_rate": 1.116695753344968e-05,
|
| 676 |
+
"loss": 0.0693,
|
| 677 |
+
"step": 4250
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.4257294429708223,
|
| 681 |
+
"grad_norm": 0.8974321484565735,
|
| 682 |
+
"learning_rate": 1.105061082024433e-05,
|
| 683 |
+
"loss": 0.0898,
|
| 684 |
+
"step": 4300
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.4423076923076923,
|
| 688 |
+
"grad_norm": 12.125022888183594,
|
| 689 |
+
"learning_rate": 1.0934264107038977e-05,
|
| 690 |
+
"loss": 0.0251,
|
| 691 |
+
"step": 4350
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 1.4588859416445623,
|
| 695 |
+
"grad_norm": 25.702117919921875,
|
| 696 |
+
"learning_rate": 1.0817917393833625e-05,
|
| 697 |
+
"loss": 0.0526,
|
| 698 |
+
"step": 4400
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.4754641909814323,
|
| 702 |
+
"grad_norm": 0.011364310048520565,
|
| 703 |
+
"learning_rate": 1.0701570680628272e-05,
|
| 704 |
+
"loss": 0.0376,
|
| 705 |
+
"step": 4450
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 1.4920424403183024,
|
| 709 |
+
"grad_norm": 0.02076265960931778,
|
| 710 |
+
"learning_rate": 1.0585223967422922e-05,
|
| 711 |
+
"loss": 0.0352,
|
| 712 |
+
"step": 4500
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 1.4920424403183024,
|
| 716 |
+
"eval_accuracy": 0.9708661417322835,
|
| 717 |
+
"eval_loss": 0.14908622205257416,
|
| 718 |
+
"eval_runtime": 8.3245,
|
| 719 |
+
"eval_samples_per_second": 305.125,
|
| 720 |
+
"eval_steps_per_second": 9.61,
|
| 721 |
+
"step": 4500
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 1.5086206896551724,
|
| 725 |
+
"grad_norm": 0.32766398787498474,
|
| 726 |
+
"learning_rate": 1.0468877254217569e-05,
|
| 727 |
+
"loss": 0.0315,
|
| 728 |
+
"step": 4550
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.5251989389920424,
|
| 732 |
+
"grad_norm": 1.0165653228759766,
|
| 733 |
+
"learning_rate": 1.0352530541012219e-05,
|
| 734 |
+
"loss": 0.0792,
|
| 735 |
+
"step": 4600
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 1.5417771883289124,
|
| 739 |
+
"grad_norm": 19.78496551513672,
|
| 740 |
+
"learning_rate": 1.0236183827806865e-05,
|
| 741 |
+
"loss": 0.0906,
|
| 742 |
+
"step": 4650
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 1.5583554376657824,
|
| 746 |
+
"grad_norm": 0.024883978068828583,
|
| 747 |
+
"learning_rate": 1.0119837114601513e-05,
|
| 748 |
+
"loss": 0.0493,
|
| 749 |
+
"step": 4700
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 1.5749336870026527,
|
| 753 |
+
"grad_norm": 0.23417970538139343,
|
| 754 |
+
"learning_rate": 1.000349040139616e-05,
|
| 755 |
+
"loss": 0.0657,
|
| 756 |
+
"step": 4750
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 1.5915119363395225,
|
| 760 |
+
"grad_norm": 0.09522684663534164,
|
| 761 |
+
"learning_rate": 9.887143688190808e-06,
|
| 762 |
+
"loss": 0.0412,
|
| 763 |
+
"step": 4800
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 1.6080901856763927,
|
| 767 |
+
"grad_norm": 0.16693036258220673,
|
| 768 |
+
"learning_rate": 9.770796974985457e-06,
|
| 769 |
+
"loss": 0.0619,
|
| 770 |
+
"step": 4850
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 1.6246684350132625,
|
| 774 |
+
"grad_norm": 25.931489944458008,
|
| 775 |
+
"learning_rate": 9.654450261780105e-06,
|
| 776 |
+
"loss": 0.0221,
|
| 777 |
+
"step": 4900
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 1.6412466843501328,
|
| 781 |
+
"grad_norm": 0.02511373534798622,
|
| 782 |
+
"learning_rate": 9.538103548574753e-06,
|
| 783 |
+
"loss": 0.0228,
|
| 784 |
+
"step": 4950
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 1.6578249336870026,
|
| 788 |
+
"grad_norm": 0.014305735006928444,
|
| 789 |
+
"learning_rate": 9.421756835369402e-06,
|
| 790 |
+
"loss": 0.0483,
|
| 791 |
+
"step": 5000
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 1.6578249336870026,
|
| 795 |
+
"eval_accuracy": 0.9692913385826771,
|
| 796 |
+
"eval_loss": 0.1490258425474167,
|
| 797 |
+
"eval_runtime": 8.5307,
|
| 798 |
+
"eval_samples_per_second": 297.75,
|
| 799 |
+
"eval_steps_per_second": 9.378,
|
| 800 |
+
"step": 5000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.6744031830238728,
|
| 804 |
+
"grad_norm": 1.0724488496780396,
|
| 805 |
+
"learning_rate": 9.30541012216405e-06,
|
| 806 |
+
"loss": 0.0485,
|
| 807 |
+
"step": 5050
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.6909814323607426,
|
| 811 |
+
"grad_norm": 0.2601702809333801,
|
| 812 |
+
"learning_rate": 9.189063408958697e-06,
|
| 813 |
+
"loss": 0.0585,
|
| 814 |
+
"step": 5100
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.7075596816976129,
|
| 818 |
+
"grad_norm": 0.4412553608417511,
|
| 819 |
+
"learning_rate": 9.072716695753345e-06,
|
| 820 |
+
"loss": 0.0278,
|
| 821 |
+
"step": 5150
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.7241379310344827,
|
| 825 |
+
"grad_norm": 20.371938705444336,
|
| 826 |
+
"learning_rate": 8.956369982547993e-06,
|
| 827 |
+
"loss": 0.0605,
|
| 828 |
+
"step": 5200
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.740716180371353,
|
| 832 |
+
"grad_norm": 33.141326904296875,
|
| 833 |
+
"learning_rate": 8.840023269342641e-06,
|
| 834 |
+
"loss": 0.0902,
|
| 835 |
+
"step": 5250
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.7572944297082227,
|
| 839 |
+
"grad_norm": 0.3250061571598053,
|
| 840 |
+
"learning_rate": 8.72367655613729e-06,
|
| 841 |
+
"loss": 0.0552,
|
| 842 |
+
"step": 5300
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.773872679045093,
|
| 846 |
+
"grad_norm": 26.73221206665039,
|
| 847 |
+
"learning_rate": 8.607329842931938e-06,
|
| 848 |
+
"loss": 0.0696,
|
| 849 |
+
"step": 5350
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.7904509283819627,
|
| 853 |
+
"grad_norm": 0.10050135850906372,
|
| 854 |
+
"learning_rate": 8.490983129726585e-06,
|
| 855 |
+
"loss": 0.07,
|
| 856 |
+
"step": 5400
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.807029177718833,
|
| 860 |
+
"grad_norm": 0.03302587568759918,
|
| 861 |
+
"learning_rate": 8.374636416521233e-06,
|
| 862 |
+
"loss": 0.0495,
|
| 863 |
+
"step": 5450
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.8236074270557028,
|
| 867 |
+
"grad_norm": 9.589860916137695,
|
| 868 |
+
"learning_rate": 8.258289703315881e-06,
|
| 869 |
+
"loss": 0.0459,
|
| 870 |
+
"step": 5500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.8236074270557028,
|
| 874 |
+
"eval_accuracy": 0.9748031496062992,
|
| 875 |
+
"eval_loss": 0.12067966908216476,
|
| 876 |
+
"eval_runtime": 8.3606,
|
| 877 |
+
"eval_samples_per_second": 303.807,
|
| 878 |
+
"eval_steps_per_second": 9.569,
|
| 879 |
+
"step": 5500
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.840185676392573,
|
| 883 |
+
"grad_norm": 21.569997787475586,
|
| 884 |
+
"learning_rate": 8.14194299011053e-06,
|
| 885 |
+
"loss": 0.0545,
|
| 886 |
+
"step": 5550
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.8567639257294428,
|
| 890 |
+
"grad_norm": 0.021630477160215378,
|
| 891 |
+
"learning_rate": 8.025596276905178e-06,
|
| 892 |
+
"loss": 0.0433,
|
| 893 |
+
"step": 5600
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.873342175066313,
|
| 897 |
+
"grad_norm": 0.01602640002965927,
|
| 898 |
+
"learning_rate": 7.909249563699826e-06,
|
| 899 |
+
"loss": 0.0281,
|
| 900 |
+
"step": 5650
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.8899204244031829,
|
| 904 |
+
"grad_norm": 0.2678263485431671,
|
| 905 |
+
"learning_rate": 7.792902850494474e-06,
|
| 906 |
+
"loss": 0.0476,
|
| 907 |
+
"step": 5700
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.906498673740053,
|
| 911 |
+
"grad_norm": 0.037133533507585526,
|
| 912 |
+
"learning_rate": 7.676556137289121e-06,
|
| 913 |
+
"loss": 0.0491,
|
| 914 |
+
"step": 5750
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.9230769230769231,
|
| 918 |
+
"grad_norm": 0.03356484696269035,
|
| 919 |
+
"learning_rate": 7.560209424083769e-06,
|
| 920 |
+
"loss": 0.0448,
|
| 921 |
+
"step": 5800
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.9396551724137931,
|
| 925 |
+
"grad_norm": 150.93212890625,
|
| 926 |
+
"learning_rate": 7.443862710878418e-06,
|
| 927 |
+
"loss": 0.0306,
|
| 928 |
+
"step": 5850
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.9562334217506632,
|
| 932 |
+
"grad_norm": 0.283079594373703,
|
| 933 |
+
"learning_rate": 7.327515997673066e-06,
|
| 934 |
+
"loss": 0.0707,
|
| 935 |
+
"step": 5900
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.9728116710875332,
|
| 939 |
+
"grad_norm": 8.607619285583496,
|
| 940 |
+
"learning_rate": 7.211169284467714e-06,
|
| 941 |
+
"loss": 0.0648,
|
| 942 |
+
"step": 5950
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.9893899204244032,
|
| 946 |
+
"grad_norm": 2.3767149448394775,
|
| 947 |
+
"learning_rate": 7.094822571262362e-06,
|
| 948 |
+
"loss": 0.0441,
|
| 949 |
+
"step": 6000
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.9893899204244032,
|
| 953 |
+
"eval_accuracy": 0.971259842519685,
|
| 954 |
+
"eval_loss": 0.13274061679840088,
|
| 955 |
+
"eval_runtime": 8.2266,
|
| 956 |
+
"eval_samples_per_second": 308.755,
|
| 957 |
+
"eval_steps_per_second": 9.725,
|
| 958 |
+
"step": 6000
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 2.0059681697612732,
|
| 962 |
+
"grad_norm": 0.05666418746113777,
|
| 963 |
+
"learning_rate": 6.97847585805701e-06,
|
| 964 |
+
"loss": 0.0261,
|
| 965 |
+
"step": 6050
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 2.022546419098143,
|
| 969 |
+
"grad_norm": 0.046994440257549286,
|
| 970 |
+
"learning_rate": 6.862129144851658e-06,
|
| 971 |
+
"loss": 0.0163,
|
| 972 |
+
"step": 6100
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.0391246684350133,
|
| 976 |
+
"grad_norm": 92.52049255371094,
|
| 977 |
+
"learning_rate": 6.745782431646306e-06,
|
| 978 |
+
"loss": 0.0156,
|
| 979 |
+
"step": 6150
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 2.055702917771883,
|
| 983 |
+
"grad_norm": 0.050654154270887375,
|
| 984 |
+
"learning_rate": 6.629435718440954e-06,
|
| 985 |
+
"loss": 0.0159,
|
| 986 |
+
"step": 6200
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.0722811671087533,
|
| 990 |
+
"grad_norm": 0.07930738478899002,
|
| 991 |
+
"learning_rate": 6.513089005235602e-06,
|
| 992 |
+
"loss": 0.0155,
|
| 993 |
+
"step": 6250
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 2.088859416445623,
|
| 997 |
+
"grad_norm": 0.006583095528185368,
|
| 998 |
+
"learning_rate": 6.39674229203025e-06,
|
| 999 |
+
"loss": 0.0122,
|
| 1000 |
+
"step": 6300
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.1054376657824934,
|
| 1004 |
+
"grad_norm": 0.009297781623899937,
|
| 1005 |
+
"learning_rate": 6.280395578824898e-06,
|
| 1006 |
+
"loss": 0.0197,
|
| 1007 |
+
"step": 6350
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 2.1220159151193636,
|
| 1011 |
+
"grad_norm": 0.06135769560933113,
|
| 1012 |
+
"learning_rate": 6.1640488656195465e-06,
|
| 1013 |
+
"loss": 0.01,
|
| 1014 |
+
"step": 6400
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.1385941644562334,
|
| 1018 |
+
"grad_norm": 0.01910097897052765,
|
| 1019 |
+
"learning_rate": 6.047702152414194e-06,
|
| 1020 |
+
"loss": 0.0166,
|
| 1021 |
+
"step": 6450
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.1551724137931036,
|
| 1025 |
+
"grad_norm": 0.009077299386262894,
|
| 1026 |
+
"learning_rate": 5.931355439208842e-06,
|
| 1027 |
+
"loss": 0.0123,
|
| 1028 |
+
"step": 6500
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.1551724137931036,
|
| 1032 |
+
"eval_accuracy": 0.9751968503937007,
|
| 1033 |
+
"eval_loss": 0.1399421989917755,
|
| 1034 |
+
"eval_runtime": 8.3995,
|
| 1035 |
+
"eval_samples_per_second": 302.399,
|
| 1036 |
+
"eval_steps_per_second": 9.524,
|
| 1037 |
+
"step": 6500
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 2.1717506631299734,
|
| 1041 |
+
"grad_norm": 0.005973361898213625,
|
| 1042 |
+
"learning_rate": 5.8150087260034905e-06,
|
| 1043 |
+
"loss": 0.006,
|
| 1044 |
+
"step": 6550
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 2.1883289124668437,
|
| 1048 |
+
"grad_norm": 0.1609232872724533,
|
| 1049 |
+
"learning_rate": 5.698662012798139e-06,
|
| 1050 |
+
"loss": 0.0244,
|
| 1051 |
+
"step": 6600
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 2.2049071618037135,
|
| 1055 |
+
"grad_norm": 30.786165237426758,
|
| 1056 |
+
"learning_rate": 5.582315299592786e-06,
|
| 1057 |
+
"loss": 0.036,
|
| 1058 |
+
"step": 6650
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 2.2214854111405837,
|
| 1062 |
+
"grad_norm": 0.007496810518205166,
|
| 1063 |
+
"learning_rate": 5.465968586387435e-06,
|
| 1064 |
+
"loss": 0.0053,
|
| 1065 |
+
"step": 6700
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 2.2380636604774535,
|
| 1069 |
+
"grad_norm": 86.89860534667969,
|
| 1070 |
+
"learning_rate": 5.349621873182083e-06,
|
| 1071 |
+
"loss": 0.021,
|
| 1072 |
+
"step": 6750
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 2.2546419098143238,
|
| 1076 |
+
"grad_norm": 0.007081742864102125,
|
| 1077 |
+
"learning_rate": 5.23327515997673e-06,
|
| 1078 |
+
"loss": 0.0184,
|
| 1079 |
+
"step": 6800
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 2.2712201591511936,
|
| 1083 |
+
"grad_norm": 11.790361404418945,
|
| 1084 |
+
"learning_rate": 5.116928446771379e-06,
|
| 1085 |
+
"loss": 0.0149,
|
| 1086 |
+
"step": 6850
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 2.287798408488064,
|
| 1090 |
+
"grad_norm": 0.012445084750652313,
|
| 1091 |
+
"learning_rate": 5.000581733566027e-06,
|
| 1092 |
+
"loss": 0.0187,
|
| 1093 |
+
"step": 6900
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 2.3043766578249336,
|
| 1097 |
+
"grad_norm": 0.00747014069929719,
|
| 1098 |
+
"learning_rate": 4.884235020360675e-06,
|
| 1099 |
+
"loss": 0.0003,
|
| 1100 |
+
"step": 6950
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 2.320954907161804,
|
| 1104 |
+
"grad_norm": 0.016270918771624565,
|
| 1105 |
+
"learning_rate": 4.7678883071553236e-06,
|
| 1106 |
+
"loss": 0.0208,
|
| 1107 |
+
"step": 7000
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 2.320954907161804,
|
| 1111 |
+
"eval_accuracy": 0.9755905511811024,
|
| 1112 |
+
"eval_loss": 0.1440856009721756,
|
| 1113 |
+
"eval_runtime": 8.3702,
|
| 1114 |
+
"eval_samples_per_second": 303.457,
|
| 1115 |
+
"eval_steps_per_second": 9.558,
|
| 1116 |
+
"step": 7000
|
| 1117 |
+
}
|
| 1118 |
+
],
|
| 1119 |
+
"logging_steps": 50,
|
| 1120 |
+
"max_steps": 9048,
|
| 1121 |
+
"num_input_tokens_seen": 0,
|
| 1122 |
+
"num_train_epochs": 3,
|
| 1123 |
+
"save_steps": 500,
|
| 1124 |
+
"stateful_callbacks": {
|
| 1125 |
+
"TrainerControl": {
|
| 1126 |
+
"args": {
|
| 1127 |
+
"should_epoch_stop": false,
|
| 1128 |
+
"should_evaluate": false,
|
| 1129 |
+
"should_log": false,
|
| 1130 |
+
"should_save": true,
|
| 1131 |
+
"should_training_stop": false
|
| 1132 |
+
},
|
| 1133 |
+
"attributes": {}
|
| 1134 |
+
}
|
| 1135 |
+
},
|
| 1136 |
+
"total_flos": 2.5784742817640743e+19,
|
| 1137 |
+
"train_batch_size": 16,
|
| 1138 |
+
"trial_name": null,
|
| 1139 |
+
"trial_params": null
|
| 1140 |
+
}
|
checkpoint-7000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:11bd20e069855e04cc796ab077650d16f81623411401c5651d4254bc705bc8e7
|
| 3 |
+
size 5777
|
checkpoint-9048/config.json
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_num_labels": 2,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Swinv2ForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"depths": [
|
| 8 |
+
2,
|
| 9 |
+
2,
|
| 10 |
+
18,
|
| 11 |
+
2
|
| 12 |
+
],
|
| 13 |
+
"drop_path_rate": 0.1,
|
| 14 |
+
"dtype": "float32",
|
| 15 |
+
"embed_dim": 192,
|
| 16 |
+
"encoder_stride": 32,
|
| 17 |
+
"hidden_act": "gelu",
|
| 18 |
+
"hidden_dropout_prob": 0.0,
|
| 19 |
+
"hidden_size": 1536,
|
| 20 |
+
"id2label": {
|
| 21 |
+
"0": "artificial",
|
| 22 |
+
"1": "real"
|
| 23 |
+
},
|
| 24 |
+
"image_size": 256,
|
| 25 |
+
"initializer_range": 0.02,
|
| 26 |
+
"label2id": {
|
| 27 |
+
"artificial": 0,
|
| 28 |
+
"real": 1
|
| 29 |
+
},
|
| 30 |
+
"layer_norm_eps": 1e-05,
|
| 31 |
+
"mlp_ratio": 4.0,
|
| 32 |
+
"model_type": "swinv2",
|
| 33 |
+
"num_channels": 3,
|
| 34 |
+
"num_heads": [
|
| 35 |
+
6,
|
| 36 |
+
12,
|
| 37 |
+
24,
|
| 38 |
+
48
|
| 39 |
+
],
|
| 40 |
+
"num_layers": 4,
|
| 41 |
+
"out_features": [
|
| 42 |
+
"stage4"
|
| 43 |
+
],
|
| 44 |
+
"out_indices": [
|
| 45 |
+
4
|
| 46 |
+
],
|
| 47 |
+
"patch_size": 4,
|
| 48 |
+
"path_norm": true,
|
| 49 |
+
"pretrained_window_sizes": [
|
| 50 |
+
12,
|
| 51 |
+
12,
|
| 52 |
+
12,
|
| 53 |
+
6
|
| 54 |
+
],
|
| 55 |
+
"problem_type": "single_label_classification",
|
| 56 |
+
"qkv_bias": true,
|
| 57 |
+
"stage_names": [
|
| 58 |
+
"stem",
|
| 59 |
+
"stage1",
|
| 60 |
+
"stage2",
|
| 61 |
+
"stage3",
|
| 62 |
+
"stage4"
|
| 63 |
+
],
|
| 64 |
+
"transformers_version": "4.57.6",
|
| 65 |
+
"use_absolute_embeddings": false,
|
| 66 |
+
"window_size": 16
|
| 67 |
+
}
|
checkpoint-9048/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e86671eb3e4a5c2d9698e16415a3e0151b54eb56efdacd50fa93906cdbb58c5b
|
| 3 |
+
size 780886280
|
checkpoint-9048/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:683b4605eef9b9f50b93d0d0c09d7e85a12d8baff6f219fd36198e5ca0adfa9a
|
| 3 |
+
size 1562050030
|
checkpoint-9048/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b0cf428e9cbf157afffbe5d57d79ec6f90df97ef4128afce405d63dd5f64b58
|
| 3 |
+
size 14645
|
checkpoint-9048/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a90e766cc9dddcb8ffc418b352672ee5843d17ac3aa7023f86221f3184afe7c1
|
| 3 |
+
size 1383
|
checkpoint-9048/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d93dac9567190ed74b2ebb3f1a4f4031f7ebfca102fa0c236bd82e8dec6a9ed
|
| 3 |
+
size 1465
|
checkpoint-9048/trainer_state.json
ADDED
|
@@ -0,0 +1,1456 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 7000,
|
| 3 |
+
"best_metric": 0.9755905511811024,
|
| 4 |
+
"best_model_checkpoint": "output/checkpoint-7000",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 9048,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.016578249336870028,
|
| 14 |
+
"grad_norm": 15.383491516113281,
|
| 15 |
+
"learning_rate": 2.1633554083885212e-06,
|
| 16 |
+
"loss": 1.0204,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.033156498673740056,
|
| 21 |
+
"grad_norm": 15.810857772827148,
|
| 22 |
+
"learning_rate": 4.370860927152319e-06,
|
| 23 |
+
"loss": 0.8263,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.04973474801061008,
|
| 28 |
+
"grad_norm": 29.156631469726562,
|
| 29 |
+
"learning_rate": 6.578366445916116e-06,
|
| 30 |
+
"loss": 0.7179,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.06631299734748011,
|
| 35 |
+
"grad_norm": 13.032061576843262,
|
| 36 |
+
"learning_rate": 8.785871964679912e-06,
|
| 37 |
+
"loss": 0.4882,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.08289124668435013,
|
| 42 |
+
"grad_norm": 7.32506799697876,
|
| 43 |
+
"learning_rate": 1.099337748344371e-05,
|
| 44 |
+
"loss": 0.4156,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.09946949602122016,
|
| 49 |
+
"grad_norm": 16.674861907958984,
|
| 50 |
+
"learning_rate": 1.3200883002207508e-05,
|
| 51 |
+
"loss": 0.3548,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.11604774535809019,
|
| 56 |
+
"grad_norm": 39.721839904785156,
|
| 57 |
+
"learning_rate": 1.5408388520971304e-05,
|
| 58 |
+
"loss": 0.3145,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.13262599469496023,
|
| 63 |
+
"grad_norm": 9.635961532592773,
|
| 64 |
+
"learning_rate": 1.76158940397351e-05,
|
| 65 |
+
"loss": 0.292,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.14920424403183025,
|
| 70 |
+
"grad_norm": 3.8946573734283447,
|
| 71 |
+
"learning_rate": 1.9823399558498897e-05,
|
| 72 |
+
"loss": 0.3039,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.16578249336870027,
|
| 77 |
+
"grad_norm": 2.919046401977539,
|
| 78 |
+
"learning_rate": 1.989296102385108e-05,
|
| 79 |
+
"loss": 0.2764,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16578249336870027,
|
| 84 |
+
"eval_accuracy": 0.9015748031496063,
|
| 85 |
+
"eval_loss": 0.25162026286125183,
|
| 86 |
+
"eval_runtime": 8.6404,
|
| 87 |
+
"eval_samples_per_second": 293.967,
|
| 88 |
+
"eval_steps_per_second": 9.259,
|
| 89 |
+
"step": 500
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.1823607427055703,
|
| 93 |
+
"grad_norm": 4.6071295738220215,
|
| 94 |
+
"learning_rate": 1.9776614310645726e-05,
|
| 95 |
+
"loss": 0.2176,
|
| 96 |
+
"step": 550
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.1989389920424403,
|
| 100 |
+
"grad_norm": 1.5469706058502197,
|
| 101 |
+
"learning_rate": 1.9660267597440373e-05,
|
| 102 |
+
"loss": 0.2785,
|
| 103 |
+
"step": 600
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.21551724137931033,
|
| 107 |
+
"grad_norm": 18.289094924926758,
|
| 108 |
+
"learning_rate": 1.9543920884235023e-05,
|
| 109 |
+
"loss": 0.1978,
|
| 110 |
+
"step": 650
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.23209549071618038,
|
| 114 |
+
"grad_norm": 16.161046981811523,
|
| 115 |
+
"learning_rate": 1.942757417102967e-05,
|
| 116 |
+
"loss": 0.2398,
|
| 117 |
+
"step": 700
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.2486737400530504,
|
| 121 |
+
"grad_norm": 4.511693000793457,
|
| 122 |
+
"learning_rate": 1.9311227457824316e-05,
|
| 123 |
+
"loss": 0.1964,
|
| 124 |
+
"step": 750
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.26525198938992045,
|
| 128 |
+
"grad_norm": 0.6140362620353699,
|
| 129 |
+
"learning_rate": 1.9194880744618966e-05,
|
| 130 |
+
"loss": 0.1817,
|
| 131 |
+
"step": 800
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.28183023872679047,
|
| 135 |
+
"grad_norm": 34.74034881591797,
|
| 136 |
+
"learning_rate": 1.9078534031413613e-05,
|
| 137 |
+
"loss": 0.1988,
|
| 138 |
+
"step": 850
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.2984084880636605,
|
| 142 |
+
"grad_norm": 2.9800262451171875,
|
| 143 |
+
"learning_rate": 1.8962187318208263e-05,
|
| 144 |
+
"loss": 0.1641,
|
| 145 |
+
"step": 900
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.3149867374005305,
|
| 149 |
+
"grad_norm": 15.570615768432617,
|
| 150 |
+
"learning_rate": 1.884584060500291e-05,
|
| 151 |
+
"loss": 0.1554,
|
| 152 |
+
"step": 950
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.33156498673740054,
|
| 156 |
+
"grad_norm": 1.9845480918884277,
|
| 157 |
+
"learning_rate": 1.872949389179756e-05,
|
| 158 |
+
"loss": 0.1855,
|
| 159 |
+
"step": 1000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.33156498673740054,
|
| 163 |
+
"eval_accuracy": 0.9358267716535433,
|
| 164 |
+
"eval_loss": 0.18621081113815308,
|
| 165 |
+
"eval_runtime": 8.3497,
|
| 166 |
+
"eval_samples_per_second": 304.203,
|
| 167 |
+
"eval_steps_per_second": 9.581,
|
| 168 |
+
"step": 1000
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.34814323607427056,
|
| 172 |
+
"grad_norm": 14.85474681854248,
|
| 173 |
+
"learning_rate": 1.8613147178592206e-05,
|
| 174 |
+
"loss": 0.1725,
|
| 175 |
+
"step": 1050
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.3647214854111406,
|
| 179 |
+
"grad_norm": 4.149899959564209,
|
| 180 |
+
"learning_rate": 1.8496800465386856e-05,
|
| 181 |
+
"loss": 0.1513,
|
| 182 |
+
"step": 1100
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.3812997347480106,
|
| 186 |
+
"grad_norm": 22.695873260498047,
|
| 187 |
+
"learning_rate": 1.8380453752181503e-05,
|
| 188 |
+
"loss": 0.1767,
|
| 189 |
+
"step": 1150
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.3978779840848806,
|
| 193 |
+
"grad_norm": 0.17689716815948486,
|
| 194 |
+
"learning_rate": 1.8264107038976153e-05,
|
| 195 |
+
"loss": 0.1623,
|
| 196 |
+
"step": 1200
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.41445623342175064,
|
| 200 |
+
"grad_norm": 30.613101959228516,
|
| 201 |
+
"learning_rate": 1.81477603257708e-05,
|
| 202 |
+
"loss": 0.1682,
|
| 203 |
+
"step": 1250
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.43103448275862066,
|
| 207 |
+
"grad_norm": 16.982980728149414,
|
| 208 |
+
"learning_rate": 1.8031413612565446e-05,
|
| 209 |
+
"loss": 0.1508,
|
| 210 |
+
"step": 1300
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.44761273209549074,
|
| 214 |
+
"grad_norm": 0.1425020545721054,
|
| 215 |
+
"learning_rate": 1.7915066899360092e-05,
|
| 216 |
+
"loss": 0.1679,
|
| 217 |
+
"step": 1350
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.46419098143236076,
|
| 221 |
+
"grad_norm": 5.924993991851807,
|
| 222 |
+
"learning_rate": 1.7798720186154742e-05,
|
| 223 |
+
"loss": 0.1387,
|
| 224 |
+
"step": 1400
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.4807692307692308,
|
| 228 |
+
"grad_norm": 5.42221212387085,
|
| 229 |
+
"learning_rate": 1.768237347294939e-05,
|
| 230 |
+
"loss": 0.1627,
|
| 231 |
+
"step": 1450
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.4973474801061008,
|
| 235 |
+
"grad_norm": 31.555957794189453,
|
| 236 |
+
"learning_rate": 1.756602675974404e-05,
|
| 237 |
+
"loss": 0.1188,
|
| 238 |
+
"step": 1500
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.4973474801061008,
|
| 242 |
+
"eval_accuracy": 0.9417322834645669,
|
| 243 |
+
"eval_loss": 0.1927609145641327,
|
| 244 |
+
"eval_runtime": 8.4174,
|
| 245 |
+
"eval_samples_per_second": 301.756,
|
| 246 |
+
"eval_steps_per_second": 9.504,
|
| 247 |
+
"step": 1500
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5139257294429708,
|
| 251 |
+
"grad_norm": 15.842325210571289,
|
| 252 |
+
"learning_rate": 1.7449680046538686e-05,
|
| 253 |
+
"loss": 0.1678,
|
| 254 |
+
"step": 1550
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5305039787798409,
|
| 258 |
+
"grad_norm": 6.040674686431885,
|
| 259 |
+
"learning_rate": 1.7333333333333336e-05,
|
| 260 |
+
"loss": 0.1215,
|
| 261 |
+
"step": 1600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.5470822281167109,
|
| 265 |
+
"grad_norm": 0.5153898596763611,
|
| 266 |
+
"learning_rate": 1.7216986620127982e-05,
|
| 267 |
+
"loss": 0.1507,
|
| 268 |
+
"step": 1650
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.5636604774535809,
|
| 272 |
+
"grad_norm": 18.98447608947754,
|
| 273 |
+
"learning_rate": 1.7100639906922632e-05,
|
| 274 |
+
"loss": 0.1481,
|
| 275 |
+
"step": 1700
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.580238726790451,
|
| 279 |
+
"grad_norm": 43.290855407714844,
|
| 280 |
+
"learning_rate": 1.698429319371728e-05,
|
| 281 |
+
"loss": 0.1422,
|
| 282 |
+
"step": 1750
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.596816976127321,
|
| 286 |
+
"grad_norm": 0.14761729538440704,
|
| 287 |
+
"learning_rate": 1.686794648051193e-05,
|
| 288 |
+
"loss": 0.1551,
|
| 289 |
+
"step": 1800
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.613395225464191,
|
| 293 |
+
"grad_norm": 96.18028259277344,
|
| 294 |
+
"learning_rate": 1.6751599767306575e-05,
|
| 295 |
+
"loss": 0.1303,
|
| 296 |
+
"step": 1850
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.629973474801061,
|
| 300 |
+
"grad_norm": 0.17378291487693787,
|
| 301 |
+
"learning_rate": 1.6635253054101222e-05,
|
| 302 |
+
"loss": 0.1484,
|
| 303 |
+
"step": 1900
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.646551724137931,
|
| 307 |
+
"grad_norm": 0.2519828677177429,
|
| 308 |
+
"learning_rate": 1.6518906340895872e-05,
|
| 309 |
+
"loss": 0.1845,
|
| 310 |
+
"step": 1950
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.6631299734748011,
|
| 314 |
+
"grad_norm": 37.454341888427734,
|
| 315 |
+
"learning_rate": 1.640255962769052e-05,
|
| 316 |
+
"loss": 0.1544,
|
| 317 |
+
"step": 2000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.6631299734748011,
|
| 321 |
+
"eval_accuracy": 0.9551181102362205,
|
| 322 |
+
"eval_loss": 0.14238852262496948,
|
| 323 |
+
"eval_runtime": 8.4401,
|
| 324 |
+
"eval_samples_per_second": 300.943,
|
| 325 |
+
"eval_steps_per_second": 9.479,
|
| 326 |
+
"step": 2000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.6797082228116711,
|
| 330 |
+
"grad_norm": 9.755090713500977,
|
| 331 |
+
"learning_rate": 1.6286212914485165e-05,
|
| 332 |
+
"loss": 0.1176,
|
| 333 |
+
"step": 2050
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.6962864721485411,
|
| 337 |
+
"grad_norm": 0.9808468222618103,
|
| 338 |
+
"learning_rate": 1.6169866201279815e-05,
|
| 339 |
+
"loss": 0.1234,
|
| 340 |
+
"step": 2100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.7128647214854111,
|
| 344 |
+
"grad_norm": 4.4084649085998535,
|
| 345 |
+
"learning_rate": 1.6053519488074462e-05,
|
| 346 |
+
"loss": 0.0882,
|
| 347 |
+
"step": 2150
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.7294429708222812,
|
| 351 |
+
"grad_norm": 35.67166519165039,
|
| 352 |
+
"learning_rate": 1.5937172774869112e-05,
|
| 353 |
+
"loss": 0.1246,
|
| 354 |
+
"step": 2200
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.7460212201591512,
|
| 358 |
+
"grad_norm": 54.88447189331055,
|
| 359 |
+
"learning_rate": 1.582082606166376e-05,
|
| 360 |
+
"loss": 0.1504,
|
| 361 |
+
"step": 2250
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.7625994694960212,
|
| 365 |
+
"grad_norm": 0.32783272862434387,
|
| 366 |
+
"learning_rate": 1.570447934845841e-05,
|
| 367 |
+
"loss": 0.1419,
|
| 368 |
+
"step": 2300
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.7791777188328912,
|
| 372 |
+
"grad_norm": 0.9903936386108398,
|
| 373 |
+
"learning_rate": 1.5588132635253055e-05,
|
| 374 |
+
"loss": 0.1154,
|
| 375 |
+
"step": 2350
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.7957559681697612,
|
| 379 |
+
"grad_norm": 0.5903331637382507,
|
| 380 |
+
"learning_rate": 1.5471785922047705e-05,
|
| 381 |
+
"loss": 0.0817,
|
| 382 |
+
"step": 2400
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.8123342175066313,
|
| 386 |
+
"grad_norm": 6.974926471710205,
|
| 387 |
+
"learning_rate": 1.5355439208842352e-05,
|
| 388 |
+
"loss": 0.1164,
|
| 389 |
+
"step": 2450
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.8289124668435013,
|
| 393 |
+
"grad_norm": 25.492650985717773,
|
| 394 |
+
"learning_rate": 1.5239092495637e-05,
|
| 395 |
+
"loss": 0.1592,
|
| 396 |
+
"step": 2500
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.8289124668435013,
|
| 400 |
+
"eval_accuracy": 0.9641732283464567,
|
| 401 |
+
"eval_loss": 0.131427600979805,
|
| 402 |
+
"eval_runtime": 8.5238,
|
| 403 |
+
"eval_samples_per_second": 297.989,
|
| 404 |
+
"eval_steps_per_second": 9.385,
|
| 405 |
+
"step": 2500
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.8454907161803713,
|
| 409 |
+
"grad_norm": 24.847103118896484,
|
| 410 |
+
"learning_rate": 1.5122745782431647e-05,
|
| 411 |
+
"loss": 0.1119,
|
| 412 |
+
"step": 2550
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.8620689655172413,
|
| 416 |
+
"grad_norm": 2.9396579265594482,
|
| 417 |
+
"learning_rate": 1.5006399069226297e-05,
|
| 418 |
+
"loss": 0.0978,
|
| 419 |
+
"step": 2600
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.8786472148541115,
|
| 423 |
+
"grad_norm": 21.483346939086914,
|
| 424 |
+
"learning_rate": 1.4890052356020943e-05,
|
| 425 |
+
"loss": 0.1233,
|
| 426 |
+
"step": 2650
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.8952254641909815,
|
| 430 |
+
"grad_norm": 0.2358592301607132,
|
| 431 |
+
"learning_rate": 1.4773705642815592e-05,
|
| 432 |
+
"loss": 0.1582,
|
| 433 |
+
"step": 2700
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.9118037135278515,
|
| 437 |
+
"grad_norm": 0.7912116646766663,
|
| 438 |
+
"learning_rate": 1.465735892961024e-05,
|
| 439 |
+
"loss": 0.1432,
|
| 440 |
+
"step": 2750
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.9283819628647215,
|
| 444 |
+
"grad_norm": 0.2763047516345978,
|
| 445 |
+
"learning_rate": 1.4541012216404888e-05,
|
| 446 |
+
"loss": 0.0988,
|
| 447 |
+
"step": 2800
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.9449602122015915,
|
| 451 |
+
"grad_norm": 0.904279887676239,
|
| 452 |
+
"learning_rate": 1.4424665503199535e-05,
|
| 453 |
+
"loss": 0.1295,
|
| 454 |
+
"step": 2850
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 0.9615384615384616,
|
| 458 |
+
"grad_norm": 0.578948974609375,
|
| 459 |
+
"learning_rate": 1.4308318789994185e-05,
|
| 460 |
+
"loss": 0.0949,
|
| 461 |
+
"step": 2900
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.9781167108753316,
|
| 465 |
+
"grad_norm": 5.489175796508789,
|
| 466 |
+
"learning_rate": 1.4191972076788831e-05,
|
| 467 |
+
"loss": 0.1224,
|
| 468 |
+
"step": 2950
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 0.9946949602122016,
|
| 472 |
+
"grad_norm": 17.271835327148438,
|
| 473 |
+
"learning_rate": 1.4075625363583481e-05,
|
| 474 |
+
"loss": 0.1476,
|
| 475 |
+
"step": 3000
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.9946949602122016,
|
| 479 |
+
"eval_accuracy": 0.9535433070866142,
|
| 480 |
+
"eval_loss": 0.16435901820659637,
|
| 481 |
+
"eval_runtime": 8.2931,
|
| 482 |
+
"eval_samples_per_second": 306.28,
|
| 483 |
+
"eval_steps_per_second": 9.647,
|
| 484 |
+
"step": 3000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 1.0112732095490715,
|
| 488 |
+
"grad_norm": 17.650592803955078,
|
| 489 |
+
"learning_rate": 1.3959278650378128e-05,
|
| 490 |
+
"loss": 0.064,
|
| 491 |
+
"step": 3050
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 1.0278514588859415,
|
| 495 |
+
"grad_norm": 0.08039892464876175,
|
| 496 |
+
"learning_rate": 1.3842931937172776e-05,
|
| 497 |
+
"loss": 0.055,
|
| 498 |
+
"step": 3100
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 1.0444297082228116,
|
| 502 |
+
"grad_norm": 36.28068542480469,
|
| 503 |
+
"learning_rate": 1.3726585223967423e-05,
|
| 504 |
+
"loss": 0.052,
|
| 505 |
+
"step": 3150
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.0610079575596818,
|
| 509 |
+
"grad_norm": 0.07026328891515732,
|
| 510 |
+
"learning_rate": 1.3610238510762073e-05,
|
| 511 |
+
"loss": 0.0706,
|
| 512 |
+
"step": 3200
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.0775862068965518,
|
| 516 |
+
"grad_norm": 6.0042853355407715,
|
| 517 |
+
"learning_rate": 1.349389179755672e-05,
|
| 518 |
+
"loss": 0.1037,
|
| 519 |
+
"step": 3250
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.0941644562334218,
|
| 523 |
+
"grad_norm": 0.8829227089881897,
|
| 524 |
+
"learning_rate": 1.337754508435137e-05,
|
| 525 |
+
"loss": 0.0536,
|
| 526 |
+
"step": 3300
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.1107427055702919,
|
| 530 |
+
"grad_norm": 54.231544494628906,
|
| 531 |
+
"learning_rate": 1.3261198371146016e-05,
|
| 532 |
+
"loss": 0.0374,
|
| 533 |
+
"step": 3350
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 1.1273209549071619,
|
| 537 |
+
"grad_norm": 0.07569246739149094,
|
| 538 |
+
"learning_rate": 1.3144851657940664e-05,
|
| 539 |
+
"loss": 0.0738,
|
| 540 |
+
"step": 3400
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 1.143899204244032,
|
| 544 |
+
"grad_norm": 0.017937984317541122,
|
| 545 |
+
"learning_rate": 1.3028504944735311e-05,
|
| 546 |
+
"loss": 0.03,
|
| 547 |
+
"step": 3450
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 1.160477453580902,
|
| 551 |
+
"grad_norm": 9.184370040893555,
|
| 552 |
+
"learning_rate": 1.2912158231529961e-05,
|
| 553 |
+
"loss": 0.0778,
|
| 554 |
+
"step": 3500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.160477453580902,
|
| 558 |
+
"eval_accuracy": 0.9677165354330709,
|
| 559 |
+
"eval_loss": 0.1568383127450943,
|
| 560 |
+
"eval_runtime": 8.2686,
|
| 561 |
+
"eval_samples_per_second": 307.187,
|
| 562 |
+
"eval_steps_per_second": 9.675,
|
| 563 |
+
"step": 3500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.177055702917772,
|
| 567 |
+
"grad_norm": 31.2453556060791,
|
| 568 |
+
"learning_rate": 1.2795811518324608e-05,
|
| 569 |
+
"loss": 0.04,
|
| 570 |
+
"step": 3550
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.193633952254642,
|
| 574 |
+
"grad_norm": 0.031678102910518646,
|
| 575 |
+
"learning_rate": 1.2679464805119258e-05,
|
| 576 |
+
"loss": 0.0687,
|
| 577 |
+
"step": 3600
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.210212201591512,
|
| 581 |
+
"grad_norm": 0.20173047482967377,
|
| 582 |
+
"learning_rate": 1.2563118091913904e-05,
|
| 583 |
+
"loss": 0.092,
|
| 584 |
+
"step": 3650
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.226790450928382,
|
| 588 |
+
"grad_norm": 0.08714820444583893,
|
| 589 |
+
"learning_rate": 1.2446771378708553e-05,
|
| 590 |
+
"loss": 0.0692,
|
| 591 |
+
"step": 3700
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.243368700265252,
|
| 595 |
+
"grad_norm": 22.537704467773438,
|
| 596 |
+
"learning_rate": 1.23304246655032e-05,
|
| 597 |
+
"loss": 0.0652,
|
| 598 |
+
"step": 3750
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.259946949602122,
|
| 602 |
+
"grad_norm": 8.494444847106934,
|
| 603 |
+
"learning_rate": 1.2214077952297849e-05,
|
| 604 |
+
"loss": 0.0541,
|
| 605 |
+
"step": 3800
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.276525198938992,
|
| 609 |
+
"grad_norm": 0.4346514046192169,
|
| 610 |
+
"learning_rate": 1.2097731239092496e-05,
|
| 611 |
+
"loss": 0.055,
|
| 612 |
+
"step": 3850
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.293103448275862,
|
| 616 |
+
"grad_norm": 0.08467918634414673,
|
| 617 |
+
"learning_rate": 1.1981384525887146e-05,
|
| 618 |
+
"loss": 0.0454,
|
| 619 |
+
"step": 3900
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.3096816976127321,
|
| 623 |
+
"grad_norm": 0.10869591683149338,
|
| 624 |
+
"learning_rate": 1.1865037812681792e-05,
|
| 625 |
+
"loss": 0.0438,
|
| 626 |
+
"step": 3950
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.3262599469496021,
|
| 630 |
+
"grad_norm": 20.05512046813965,
|
| 631 |
+
"learning_rate": 1.174869109947644e-05,
|
| 632 |
+
"loss": 0.0661,
|
| 633 |
+
"step": 4000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.3262599469496021,
|
| 637 |
+
"eval_accuracy": 0.9696850393700788,
|
| 638 |
+
"eval_loss": 0.14853054285049438,
|
| 639 |
+
"eval_runtime": 8.4299,
|
| 640 |
+
"eval_samples_per_second": 301.307,
|
| 641 |
+
"eval_steps_per_second": 9.49,
|
| 642 |
+
"step": 4000
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 1.3428381962864722,
|
| 646 |
+
"grad_norm": 0.03564869612455368,
|
| 647 |
+
"learning_rate": 1.1632344386271089e-05,
|
| 648 |
+
"loss": 0.0529,
|
| 649 |
+
"step": 4050
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 1.3594164456233422,
|
| 653 |
+
"grad_norm": 0.020624179393053055,
|
| 654 |
+
"learning_rate": 1.1515997673065737e-05,
|
| 655 |
+
"loss": 0.0525,
|
| 656 |
+
"step": 4100
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.3759946949602122,
|
| 660 |
+
"grad_norm": 10.6432466506958,
|
| 661 |
+
"learning_rate": 1.1399650959860384e-05,
|
| 662 |
+
"loss": 0.0502,
|
| 663 |
+
"step": 4150
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 1.3925729442970822,
|
| 667 |
+
"grad_norm": 0.02102123387157917,
|
| 668 |
+
"learning_rate": 1.1283304246655034e-05,
|
| 669 |
+
"loss": 0.0582,
|
| 670 |
+
"step": 4200
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.4091511936339522,
|
| 674 |
+
"grad_norm": 32.820335388183594,
|
| 675 |
+
"learning_rate": 1.116695753344968e-05,
|
| 676 |
+
"loss": 0.0693,
|
| 677 |
+
"step": 4250
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 1.4257294429708223,
|
| 681 |
+
"grad_norm": 0.8974321484565735,
|
| 682 |
+
"learning_rate": 1.105061082024433e-05,
|
| 683 |
+
"loss": 0.0898,
|
| 684 |
+
"step": 4300
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.4423076923076923,
|
| 688 |
+
"grad_norm": 12.125022888183594,
|
| 689 |
+
"learning_rate": 1.0934264107038977e-05,
|
| 690 |
+
"loss": 0.0251,
|
| 691 |
+
"step": 4350
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 1.4588859416445623,
|
| 695 |
+
"grad_norm": 25.702117919921875,
|
| 696 |
+
"learning_rate": 1.0817917393833625e-05,
|
| 697 |
+
"loss": 0.0526,
|
| 698 |
+
"step": 4400
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.4754641909814323,
|
| 702 |
+
"grad_norm": 0.011364310048520565,
|
| 703 |
+
"learning_rate": 1.0701570680628272e-05,
|
| 704 |
+
"loss": 0.0376,
|
| 705 |
+
"step": 4450
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 1.4920424403183024,
|
| 709 |
+
"grad_norm": 0.02076265960931778,
|
| 710 |
+
"learning_rate": 1.0585223967422922e-05,
|
| 711 |
+
"loss": 0.0352,
|
| 712 |
+
"step": 4500
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 1.4920424403183024,
|
| 716 |
+
"eval_accuracy": 0.9708661417322835,
|
| 717 |
+
"eval_loss": 0.14908622205257416,
|
| 718 |
+
"eval_runtime": 8.3245,
|
| 719 |
+
"eval_samples_per_second": 305.125,
|
| 720 |
+
"eval_steps_per_second": 9.61,
|
| 721 |
+
"step": 4500
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 1.5086206896551724,
|
| 725 |
+
"grad_norm": 0.32766398787498474,
|
| 726 |
+
"learning_rate": 1.0468877254217569e-05,
|
| 727 |
+
"loss": 0.0315,
|
| 728 |
+
"step": 4550
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.5251989389920424,
|
| 732 |
+
"grad_norm": 1.0165653228759766,
|
| 733 |
+
"learning_rate": 1.0352530541012219e-05,
|
| 734 |
+
"loss": 0.0792,
|
| 735 |
+
"step": 4600
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 1.5417771883289124,
|
| 739 |
+
"grad_norm": 19.78496551513672,
|
| 740 |
+
"learning_rate": 1.0236183827806865e-05,
|
| 741 |
+
"loss": 0.0906,
|
| 742 |
+
"step": 4650
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 1.5583554376657824,
|
| 746 |
+
"grad_norm": 0.024883978068828583,
|
| 747 |
+
"learning_rate": 1.0119837114601513e-05,
|
| 748 |
+
"loss": 0.0493,
|
| 749 |
+
"step": 4700
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 1.5749336870026527,
|
| 753 |
+
"grad_norm": 0.23417970538139343,
|
| 754 |
+
"learning_rate": 1.000349040139616e-05,
|
| 755 |
+
"loss": 0.0657,
|
| 756 |
+
"step": 4750
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 1.5915119363395225,
|
| 760 |
+
"grad_norm": 0.09522684663534164,
|
| 761 |
+
"learning_rate": 9.887143688190808e-06,
|
| 762 |
+
"loss": 0.0412,
|
| 763 |
+
"step": 4800
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 1.6080901856763927,
|
| 767 |
+
"grad_norm": 0.16693036258220673,
|
| 768 |
+
"learning_rate": 9.770796974985457e-06,
|
| 769 |
+
"loss": 0.0619,
|
| 770 |
+
"step": 4850
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 1.6246684350132625,
|
| 774 |
+
"grad_norm": 25.931489944458008,
|
| 775 |
+
"learning_rate": 9.654450261780105e-06,
|
| 776 |
+
"loss": 0.0221,
|
| 777 |
+
"step": 4900
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 1.6412466843501328,
|
| 781 |
+
"grad_norm": 0.02511373534798622,
|
| 782 |
+
"learning_rate": 9.538103548574753e-06,
|
| 783 |
+
"loss": 0.0228,
|
| 784 |
+
"step": 4950
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 1.6578249336870026,
|
| 788 |
+
"grad_norm": 0.014305735006928444,
|
| 789 |
+
"learning_rate": 9.421756835369402e-06,
|
| 790 |
+
"loss": 0.0483,
|
| 791 |
+
"step": 5000
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 1.6578249336870026,
|
| 795 |
+
"eval_accuracy": 0.9692913385826771,
|
| 796 |
+
"eval_loss": 0.1490258425474167,
|
| 797 |
+
"eval_runtime": 8.5307,
|
| 798 |
+
"eval_samples_per_second": 297.75,
|
| 799 |
+
"eval_steps_per_second": 9.378,
|
| 800 |
+
"step": 5000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.6744031830238728,
|
| 804 |
+
"grad_norm": 1.0724488496780396,
|
| 805 |
+
"learning_rate": 9.30541012216405e-06,
|
| 806 |
+
"loss": 0.0485,
|
| 807 |
+
"step": 5050
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.6909814323607426,
|
| 811 |
+
"grad_norm": 0.2601702809333801,
|
| 812 |
+
"learning_rate": 9.189063408958697e-06,
|
| 813 |
+
"loss": 0.0585,
|
| 814 |
+
"step": 5100
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.7075596816976129,
|
| 818 |
+
"grad_norm": 0.4412553608417511,
|
| 819 |
+
"learning_rate": 9.072716695753345e-06,
|
| 820 |
+
"loss": 0.0278,
|
| 821 |
+
"step": 5150
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.7241379310344827,
|
| 825 |
+
"grad_norm": 20.371938705444336,
|
| 826 |
+
"learning_rate": 8.956369982547993e-06,
|
| 827 |
+
"loss": 0.0605,
|
| 828 |
+
"step": 5200
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.740716180371353,
|
| 832 |
+
"grad_norm": 33.141326904296875,
|
| 833 |
+
"learning_rate": 8.840023269342641e-06,
|
| 834 |
+
"loss": 0.0902,
|
| 835 |
+
"step": 5250
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.7572944297082227,
|
| 839 |
+
"grad_norm": 0.3250061571598053,
|
| 840 |
+
"learning_rate": 8.72367655613729e-06,
|
| 841 |
+
"loss": 0.0552,
|
| 842 |
+
"step": 5300
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.773872679045093,
|
| 846 |
+
"grad_norm": 26.73221206665039,
|
| 847 |
+
"learning_rate": 8.607329842931938e-06,
|
| 848 |
+
"loss": 0.0696,
|
| 849 |
+
"step": 5350
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.7904509283819627,
|
| 853 |
+
"grad_norm": 0.10050135850906372,
|
| 854 |
+
"learning_rate": 8.490983129726585e-06,
|
| 855 |
+
"loss": 0.07,
|
| 856 |
+
"step": 5400
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.807029177718833,
|
| 860 |
+
"grad_norm": 0.03302587568759918,
|
| 861 |
+
"learning_rate": 8.374636416521233e-06,
|
| 862 |
+
"loss": 0.0495,
|
| 863 |
+
"step": 5450
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.8236074270557028,
|
| 867 |
+
"grad_norm": 9.589860916137695,
|
| 868 |
+
"learning_rate": 8.258289703315881e-06,
|
| 869 |
+
"loss": 0.0459,
|
| 870 |
+
"step": 5500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.8236074270557028,
|
| 874 |
+
"eval_accuracy": 0.9748031496062992,
|
| 875 |
+
"eval_loss": 0.12067966908216476,
|
| 876 |
+
"eval_runtime": 8.3606,
|
| 877 |
+
"eval_samples_per_second": 303.807,
|
| 878 |
+
"eval_steps_per_second": 9.569,
|
| 879 |
+
"step": 5500
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.840185676392573,
|
| 883 |
+
"grad_norm": 21.569997787475586,
|
| 884 |
+
"learning_rate": 8.14194299011053e-06,
|
| 885 |
+
"loss": 0.0545,
|
| 886 |
+
"step": 5550
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.8567639257294428,
|
| 890 |
+
"grad_norm": 0.021630477160215378,
|
| 891 |
+
"learning_rate": 8.025596276905178e-06,
|
| 892 |
+
"loss": 0.0433,
|
| 893 |
+
"step": 5600
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.873342175066313,
|
| 897 |
+
"grad_norm": 0.01602640002965927,
|
| 898 |
+
"learning_rate": 7.909249563699826e-06,
|
| 899 |
+
"loss": 0.0281,
|
| 900 |
+
"step": 5650
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.8899204244031829,
|
| 904 |
+
"grad_norm": 0.2678263485431671,
|
| 905 |
+
"learning_rate": 7.792902850494474e-06,
|
| 906 |
+
"loss": 0.0476,
|
| 907 |
+
"step": 5700
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.906498673740053,
|
| 911 |
+
"grad_norm": 0.037133533507585526,
|
| 912 |
+
"learning_rate": 7.676556137289121e-06,
|
| 913 |
+
"loss": 0.0491,
|
| 914 |
+
"step": 5750
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.9230769230769231,
|
| 918 |
+
"grad_norm": 0.03356484696269035,
|
| 919 |
+
"learning_rate": 7.560209424083769e-06,
|
| 920 |
+
"loss": 0.0448,
|
| 921 |
+
"step": 5800
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.9396551724137931,
|
| 925 |
+
"grad_norm": 150.93212890625,
|
| 926 |
+
"learning_rate": 7.443862710878418e-06,
|
| 927 |
+
"loss": 0.0306,
|
| 928 |
+
"step": 5850
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.9562334217506632,
|
| 932 |
+
"grad_norm": 0.283079594373703,
|
| 933 |
+
"learning_rate": 7.327515997673066e-06,
|
| 934 |
+
"loss": 0.0707,
|
| 935 |
+
"step": 5900
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.9728116710875332,
|
| 939 |
+
"grad_norm": 8.607619285583496,
|
| 940 |
+
"learning_rate": 7.211169284467714e-06,
|
| 941 |
+
"loss": 0.0648,
|
| 942 |
+
"step": 5950
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.9893899204244032,
|
| 946 |
+
"grad_norm": 2.3767149448394775,
|
| 947 |
+
"learning_rate": 7.094822571262362e-06,
|
| 948 |
+
"loss": 0.0441,
|
| 949 |
+
"step": 6000
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.9893899204244032,
|
| 953 |
+
"eval_accuracy": 0.971259842519685,
|
| 954 |
+
"eval_loss": 0.13274061679840088,
|
| 955 |
+
"eval_runtime": 8.2266,
|
| 956 |
+
"eval_samples_per_second": 308.755,
|
| 957 |
+
"eval_steps_per_second": 9.725,
|
| 958 |
+
"step": 6000
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 2.0059681697612732,
|
| 962 |
+
"grad_norm": 0.05666418746113777,
|
| 963 |
+
"learning_rate": 6.97847585805701e-06,
|
| 964 |
+
"loss": 0.0261,
|
| 965 |
+
"step": 6050
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 2.022546419098143,
|
| 969 |
+
"grad_norm": 0.046994440257549286,
|
| 970 |
+
"learning_rate": 6.862129144851658e-06,
|
| 971 |
+
"loss": 0.0163,
|
| 972 |
+
"step": 6100
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.0391246684350133,
|
| 976 |
+
"grad_norm": 92.52049255371094,
|
| 977 |
+
"learning_rate": 6.745782431646306e-06,
|
| 978 |
+
"loss": 0.0156,
|
| 979 |
+
"step": 6150
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 2.055702917771883,
|
| 983 |
+
"grad_norm": 0.050654154270887375,
|
| 984 |
+
"learning_rate": 6.629435718440954e-06,
|
| 985 |
+
"loss": 0.0159,
|
| 986 |
+
"step": 6200
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.0722811671087533,
|
| 990 |
+
"grad_norm": 0.07930738478899002,
|
| 991 |
+
"learning_rate": 6.513089005235602e-06,
|
| 992 |
+
"loss": 0.0155,
|
| 993 |
+
"step": 6250
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 2.088859416445623,
|
| 997 |
+
"grad_norm": 0.006583095528185368,
|
| 998 |
+
"learning_rate": 6.39674229203025e-06,
|
| 999 |
+
"loss": 0.0122,
|
| 1000 |
+
"step": 6300
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.1054376657824934,
|
| 1004 |
+
"grad_norm": 0.009297781623899937,
|
| 1005 |
+
"learning_rate": 6.280395578824898e-06,
|
| 1006 |
+
"loss": 0.0197,
|
| 1007 |
+
"step": 6350
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 2.1220159151193636,
|
| 1011 |
+
"grad_norm": 0.06135769560933113,
|
| 1012 |
+
"learning_rate": 6.1640488656195465e-06,
|
| 1013 |
+
"loss": 0.01,
|
| 1014 |
+
"step": 6400
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.1385941644562334,
|
| 1018 |
+
"grad_norm": 0.01910097897052765,
|
| 1019 |
+
"learning_rate": 6.047702152414194e-06,
|
| 1020 |
+
"loss": 0.0166,
|
| 1021 |
+
"step": 6450
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.1551724137931036,
|
| 1025 |
+
"grad_norm": 0.009077299386262894,
|
| 1026 |
+
"learning_rate": 5.931355439208842e-06,
|
| 1027 |
+
"loss": 0.0123,
|
| 1028 |
+
"step": 6500
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.1551724137931036,
|
| 1032 |
+
"eval_accuracy": 0.9751968503937007,
|
| 1033 |
+
"eval_loss": 0.1399421989917755,
|
| 1034 |
+
"eval_runtime": 8.3995,
|
| 1035 |
+
"eval_samples_per_second": 302.399,
|
| 1036 |
+
"eval_steps_per_second": 9.524,
|
| 1037 |
+
"step": 6500
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 2.1717506631299734,
|
| 1041 |
+
"grad_norm": 0.005973361898213625,
|
| 1042 |
+
"learning_rate": 5.8150087260034905e-06,
|
| 1043 |
+
"loss": 0.006,
|
| 1044 |
+
"step": 6550
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 2.1883289124668437,
|
| 1048 |
+
"grad_norm": 0.1609232872724533,
|
| 1049 |
+
"learning_rate": 5.698662012798139e-06,
|
| 1050 |
+
"loss": 0.0244,
|
| 1051 |
+
"step": 6600
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 2.2049071618037135,
|
| 1055 |
+
"grad_norm": 30.786165237426758,
|
| 1056 |
+
"learning_rate": 5.582315299592786e-06,
|
| 1057 |
+
"loss": 0.036,
|
| 1058 |
+
"step": 6650
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 2.2214854111405837,
|
| 1062 |
+
"grad_norm": 0.007496810518205166,
|
| 1063 |
+
"learning_rate": 5.465968586387435e-06,
|
| 1064 |
+
"loss": 0.0053,
|
| 1065 |
+
"step": 6700
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 2.2380636604774535,
|
| 1069 |
+
"grad_norm": 86.89860534667969,
|
| 1070 |
+
"learning_rate": 5.349621873182083e-06,
|
| 1071 |
+
"loss": 0.021,
|
| 1072 |
+
"step": 6750
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 2.2546419098143238,
|
| 1076 |
+
"grad_norm": 0.007081742864102125,
|
| 1077 |
+
"learning_rate": 5.23327515997673e-06,
|
| 1078 |
+
"loss": 0.0184,
|
| 1079 |
+
"step": 6800
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 2.2712201591511936,
|
| 1083 |
+
"grad_norm": 11.790361404418945,
|
| 1084 |
+
"learning_rate": 5.116928446771379e-06,
|
| 1085 |
+
"loss": 0.0149,
|
| 1086 |
+
"step": 6850
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 2.287798408488064,
|
| 1090 |
+
"grad_norm": 0.012445084750652313,
|
| 1091 |
+
"learning_rate": 5.000581733566027e-06,
|
| 1092 |
+
"loss": 0.0187,
|
| 1093 |
+
"step": 6900
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 2.3043766578249336,
|
| 1097 |
+
"grad_norm": 0.00747014069929719,
|
| 1098 |
+
"learning_rate": 4.884235020360675e-06,
|
| 1099 |
+
"loss": 0.0003,
|
| 1100 |
+
"step": 6950
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 2.320954907161804,
|
| 1104 |
+
"grad_norm": 0.016270918771624565,
|
| 1105 |
+
"learning_rate": 4.7678883071553236e-06,
|
| 1106 |
+
"loss": 0.0208,
|
| 1107 |
+
"step": 7000
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 2.320954907161804,
|
| 1111 |
+
"eval_accuracy": 0.9755905511811024,
|
| 1112 |
+
"eval_loss": 0.1440856009721756,
|
| 1113 |
+
"eval_runtime": 8.3702,
|
| 1114 |
+
"eval_samples_per_second": 303.457,
|
| 1115 |
+
"eval_steps_per_second": 9.558,
|
| 1116 |
+
"step": 7000
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.3375331564986737,
|
| 1120 |
+
"grad_norm": 0.0076462519355118275,
|
| 1121 |
+
"learning_rate": 4.651541593949971e-06,
|
| 1122 |
+
"loss": 0.0134,
|
| 1123 |
+
"step": 7050
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.354111405835544,
|
| 1127 |
+
"grad_norm": 0.008808553218841553,
|
| 1128 |
+
"learning_rate": 4.535194880744619e-06,
|
| 1129 |
+
"loss": 0.0092,
|
| 1130 |
+
"step": 7100
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.3706896551724137,
|
| 1134 |
+
"grad_norm": 8.406987190246582,
|
| 1135 |
+
"learning_rate": 4.418848167539268e-06,
|
| 1136 |
+
"loss": 0.0368,
|
| 1137 |
+
"step": 7150
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.387267904509284,
|
| 1141 |
+
"grad_norm": 0.004396683536469936,
|
| 1142 |
+
"learning_rate": 4.302501454333916e-06,
|
| 1143 |
+
"loss": 0.0035,
|
| 1144 |
+
"step": 7200
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.4038461538461537,
|
| 1148 |
+
"grad_norm": 0.005913051310926676,
|
| 1149 |
+
"learning_rate": 4.186154741128563e-06,
|
| 1150 |
+
"loss": 0.0021,
|
| 1151 |
+
"step": 7250
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.420424403183024,
|
| 1155 |
+
"grad_norm": 0.00849180482327938,
|
| 1156 |
+
"learning_rate": 4.069808027923212e-06,
|
| 1157 |
+
"loss": 0.0135,
|
| 1158 |
+
"step": 7300
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.437002652519894,
|
| 1162 |
+
"grad_norm": 18.357879638671875,
|
| 1163 |
+
"learning_rate": 3.95346131471786e-06,
|
| 1164 |
+
"loss": 0.0119,
|
| 1165 |
+
"step": 7350
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.453580901856764,
|
| 1169 |
+
"grad_norm": 0.005445088259875774,
|
| 1170 |
+
"learning_rate": 3.8371146015125074e-06,
|
| 1171 |
+
"loss": 0.018,
|
| 1172 |
+
"step": 7400
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.470159151193634,
|
| 1176 |
+
"grad_norm": 97.43321990966797,
|
| 1177 |
+
"learning_rate": 3.7207678883071557e-06,
|
| 1178 |
+
"loss": 0.0329,
|
| 1179 |
+
"step": 7450
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.486737400530504,
|
| 1183 |
+
"grad_norm": 0.003963190596550703,
|
| 1184 |
+
"learning_rate": 3.6044211751018036e-06,
|
| 1185 |
+
"loss": 0.0385,
|
| 1186 |
+
"step": 7500
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.486737400530504,
|
| 1190 |
+
"eval_accuracy": 0.974015748031496,
|
| 1191 |
+
"eval_loss": 0.1490032821893692,
|
| 1192 |
+
"eval_runtime": 8.4191,
|
| 1193 |
+
"eval_samples_per_second": 301.694,
|
| 1194 |
+
"eval_steps_per_second": 9.502,
|
| 1195 |
+
"step": 7500
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 2.503315649867374,
|
| 1199 |
+
"grad_norm": 0.004599283449351788,
|
| 1200 |
+
"learning_rate": 3.488074461896452e-06,
|
| 1201 |
+
"loss": 0.0151,
|
| 1202 |
+
"step": 7550
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.519893899204244,
|
| 1206 |
+
"grad_norm": 0.032206956297159195,
|
| 1207 |
+
"learning_rate": 3.3717277486911e-06,
|
| 1208 |
+
"loss": 0.0075,
|
| 1209 |
+
"step": 7600
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 2.536472148541114,
|
| 1213 |
+
"grad_norm": 7.669720649719238,
|
| 1214 |
+
"learning_rate": 3.255381035485748e-06,
|
| 1215 |
+
"loss": 0.0329,
|
| 1216 |
+
"step": 7650
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 2.553050397877984,
|
| 1220 |
+
"grad_norm": 0.045368775725364685,
|
| 1221 |
+
"learning_rate": 3.139034322280396e-06,
|
| 1222 |
+
"loss": 0.0394,
|
| 1223 |
+
"step": 7700
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 2.569628647214854,
|
| 1227 |
+
"grad_norm": 0.012171960435807705,
|
| 1228 |
+
"learning_rate": 3.022687609075044e-06,
|
| 1229 |
+
"loss": 0.0035,
|
| 1230 |
+
"step": 7750
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 2.586206896551724,
|
| 1234 |
+
"grad_norm": 0.06905154883861542,
|
| 1235 |
+
"learning_rate": 2.906340895869692e-06,
|
| 1236 |
+
"loss": 0.0241,
|
| 1237 |
+
"step": 7800
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 2.602785145888594,
|
| 1241 |
+
"grad_norm": 0.00501677393913269,
|
| 1242 |
+
"learning_rate": 2.78999418266434e-06,
|
| 1243 |
+
"loss": 0.0073,
|
| 1244 |
+
"step": 7850
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 2.6193633952254642,
|
| 1248 |
+
"grad_norm": 0.04552486911416054,
|
| 1249 |
+
"learning_rate": 2.673647469458988e-06,
|
| 1250 |
+
"loss": 0.0064,
|
| 1251 |
+
"step": 7900
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 2.635941644562334,
|
| 1255 |
+
"grad_norm": 0.04517211765050888,
|
| 1256 |
+
"learning_rate": 2.5573007562536362e-06,
|
| 1257 |
+
"loss": 0.0223,
|
| 1258 |
+
"step": 7950
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 2.6525198938992043,
|
| 1262 |
+
"grad_norm": 0.008471992798149586,
|
| 1263 |
+
"learning_rate": 2.440954043048284e-06,
|
| 1264 |
+
"loss": 0.0241,
|
| 1265 |
+
"step": 8000
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 2.6525198938992043,
|
| 1269 |
+
"eval_accuracy": 0.9708661417322835,
|
| 1270 |
+
"eval_loss": 0.1622270792722702,
|
| 1271 |
+
"eval_runtime": 8.3014,
|
| 1272 |
+
"eval_samples_per_second": 305.972,
|
| 1273 |
+
"eval_steps_per_second": 9.637,
|
| 1274 |
+
"step": 8000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 2.6690981432360745,
|
| 1278 |
+
"grad_norm": 0.169709712266922,
|
| 1279 |
+
"learning_rate": 2.3246073298429324e-06,
|
| 1280 |
+
"loss": 0.0124,
|
| 1281 |
+
"step": 8050
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 2.6856763925729443,
|
| 1285 |
+
"grad_norm": 1.3567891120910645,
|
| 1286 |
+
"learning_rate": 2.2082606166375803e-06,
|
| 1287 |
+
"loss": 0.0065,
|
| 1288 |
+
"step": 8100
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 2.702254641909814,
|
| 1292 |
+
"grad_norm": 18.267772674560547,
|
| 1293 |
+
"learning_rate": 2.091913903432228e-06,
|
| 1294 |
+
"loss": 0.0129,
|
| 1295 |
+
"step": 8150
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 2.7188328912466844,
|
| 1299 |
+
"grad_norm": 0.012821214273571968,
|
| 1300 |
+
"learning_rate": 1.9755671902268765e-06,
|
| 1301 |
+
"loss": 0.0071,
|
| 1302 |
+
"step": 8200
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 2.7354111405835546,
|
| 1306 |
+
"grad_norm": 0.004050163086503744,
|
| 1307 |
+
"learning_rate": 1.8592204770215244e-06,
|
| 1308 |
+
"loss": 0.0136,
|
| 1309 |
+
"step": 8250
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 2.7519893899204244,
|
| 1313 |
+
"grad_norm": 0.008309832774102688,
|
| 1314 |
+
"learning_rate": 1.7428737638161724e-06,
|
| 1315 |
+
"loss": 0.0136,
|
| 1316 |
+
"step": 8300
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.768567639257294,
|
| 1320 |
+
"grad_norm": 0.038879744708538055,
|
| 1321 |
+
"learning_rate": 1.6265270506108205e-06,
|
| 1322 |
+
"loss": 0.0036,
|
| 1323 |
+
"step": 8350
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 2.7851458885941645,
|
| 1327 |
+
"grad_norm": 0.045335281640291214,
|
| 1328 |
+
"learning_rate": 1.5101803374054686e-06,
|
| 1329 |
+
"loss": 0.0089,
|
| 1330 |
+
"step": 8400
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 2.8017241379310347,
|
| 1334 |
+
"grad_norm": 0.09732075035572052,
|
| 1335 |
+
"learning_rate": 1.3938336242001165e-06,
|
| 1336 |
+
"loss": 0.0002,
|
| 1337 |
+
"step": 8450
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 2.8183023872679045,
|
| 1341 |
+
"grad_norm": 0.004367977846413851,
|
| 1342 |
+
"learning_rate": 1.2774869109947646e-06,
|
| 1343 |
+
"loss": 0.0112,
|
| 1344 |
+
"step": 8500
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 2.8183023872679045,
|
| 1348 |
+
"eval_accuracy": 0.9751968503937007,
|
| 1349 |
+
"eval_loss": 0.15158241987228394,
|
| 1350 |
+
"eval_runtime": 8.4462,
|
| 1351 |
+
"eval_samples_per_second": 300.725,
|
| 1352 |
+
"eval_steps_per_second": 9.472,
|
| 1353 |
+
"step": 8500
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.8348806366047743,
|
| 1357 |
+
"grad_norm": 0.003420003689825535,
|
| 1358 |
+
"learning_rate": 1.1611401977894125e-06,
|
| 1359 |
+
"loss": 0.0128,
|
| 1360 |
+
"step": 8550
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.8514588859416445,
|
| 1364 |
+
"grad_norm": 0.020186182111501694,
|
| 1365 |
+
"learning_rate": 1.0447934845840606e-06,
|
| 1366 |
+
"loss": 0.0188,
|
| 1367 |
+
"step": 8600
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.868037135278515,
|
| 1371 |
+
"grad_norm": 0.021418306976556778,
|
| 1372 |
+
"learning_rate": 9.284467713787087e-07,
|
| 1373 |
+
"loss": 0.0217,
|
| 1374 |
+
"step": 8650
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.8846153846153846,
|
| 1378 |
+
"grad_norm": 0.0053920382633805275,
|
| 1379 |
+
"learning_rate": 8.121000581733566e-07,
|
| 1380 |
+
"loss": 0.0141,
|
| 1381 |
+
"step": 8700
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.9011936339522544,
|
| 1385 |
+
"grad_norm": 0.019735030829906464,
|
| 1386 |
+
"learning_rate": 6.957533449680047e-07,
|
| 1387 |
+
"loss": 0.0268,
|
| 1388 |
+
"step": 8750
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.9177718832891246,
|
| 1392 |
+
"grad_norm": 0.027690809220075607,
|
| 1393 |
+
"learning_rate": 5.794066317626527e-07,
|
| 1394 |
+
"loss": 0.0231,
|
| 1395 |
+
"step": 8800
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 2.934350132625995,
|
| 1399 |
+
"grad_norm": 1.5282634496688843,
|
| 1400 |
+
"learning_rate": 4.6305991855730076e-07,
|
| 1401 |
+
"loss": 0.0094,
|
| 1402 |
+
"step": 8850
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.9509283819628647,
|
| 1406 |
+
"grad_norm": 0.007764532696455717,
|
| 1407 |
+
"learning_rate": 3.4671320535194885e-07,
|
| 1408 |
+
"loss": 0.0053,
|
| 1409 |
+
"step": 8900
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 2.967506631299735,
|
| 1413 |
+
"grad_norm": 0.03694835305213928,
|
| 1414 |
+
"learning_rate": 2.3036649214659686e-07,
|
| 1415 |
+
"loss": 0.0089,
|
| 1416 |
+
"step": 8950
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 2.9840848806366047,
|
| 1420 |
+
"grad_norm": 0.009050080552697182,
|
| 1421 |
+
"learning_rate": 1.1401977894124491e-07,
|
| 1422 |
+
"loss": 0.0075,
|
| 1423 |
+
"step": 9000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 2.9840848806366047,
|
| 1427 |
+
"eval_accuracy": 0.9748031496062992,
|
| 1428 |
+
"eval_loss": 0.15235309302806854,
|
| 1429 |
+
"eval_runtime": 8.4365,
|
| 1430 |
+
"eval_samples_per_second": 301.073,
|
| 1431 |
+
"eval_steps_per_second": 9.483,
|
| 1432 |
+
"step": 9000
|
| 1433 |
+
}
|
| 1434 |
+
],
|
| 1435 |
+
"logging_steps": 50,
|
| 1436 |
+
"max_steps": 9048,
|
| 1437 |
+
"num_input_tokens_seen": 0,
|
| 1438 |
+
"num_train_epochs": 3,
|
| 1439 |
+
"save_steps": 500,
|
| 1440 |
+
"stateful_callbacks": {
|
| 1441 |
+
"TrainerControl": {
|
| 1442 |
+
"args": {
|
| 1443 |
+
"should_epoch_stop": false,
|
| 1444 |
+
"should_evaluate": false,
|
| 1445 |
+
"should_log": false,
|
| 1446 |
+
"should_save": true,
|
| 1447 |
+
"should_training_stop": true
|
| 1448 |
+
},
|
| 1449 |
+
"attributes": {}
|
| 1450 |
+
}
|
| 1451 |
+
},
|
| 1452 |
+
"total_flos": 3.3327379956546404e+19,
|
| 1453 |
+
"train_batch_size": 16,
|
| 1454 |
+
"trial_name": null,
|
| 1455 |
+
"trial_params": null
|
| 1456 |
+
}
|
checkpoint-9048/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:11bd20e069855e04cc796ab077650d16f81623411401c5651d4254bc705bc8e7
|
| 3 |
+
size 5777
|
config.json
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_num_labels": 2,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Swinv2ForImageClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.0,
|
| 7 |
+
"depths": [
|
| 8 |
+
2,
|
| 9 |
+
2,
|
| 10 |
+
18,
|
| 11 |
+
2
|
| 12 |
+
],
|
| 13 |
+
"drop_path_rate": 0.1,
|
| 14 |
+
"dtype": "float32",
|
| 15 |
+
"embed_dim": 192,
|
| 16 |
+
"encoder_stride": 32,
|
| 17 |
+
"hidden_act": "gelu",
|
| 18 |
+
"hidden_dropout_prob": 0.0,
|
| 19 |
+
"hidden_size": 1536,
|
| 20 |
+
"id2label": {
|
| 21 |
+
"0": "artificial",
|
| 22 |
+
"1": "real"
|
| 23 |
+
},
|
| 24 |
+
"image_size": 256,
|
| 25 |
+
"initializer_range": 0.02,
|
| 26 |
+
"label2id": {
|
| 27 |
+
"artificial": 0,
|
| 28 |
+
"real": 1
|
| 29 |
+
},
|
| 30 |
+
"layer_norm_eps": 1e-05,
|
| 31 |
+
"mlp_ratio": 4.0,
|
| 32 |
+
"model_type": "swinv2",
|
| 33 |
+
"num_channels": 3,
|
| 34 |
+
"num_heads": [
|
| 35 |
+
6,
|
| 36 |
+
12,
|
| 37 |
+
24,
|
| 38 |
+
48
|
| 39 |
+
],
|
| 40 |
+
"num_layers": 4,
|
| 41 |
+
"out_features": [
|
| 42 |
+
"stage4"
|
| 43 |
+
],
|
| 44 |
+
"out_indices": [
|
| 45 |
+
4
|
| 46 |
+
],
|
| 47 |
+
"patch_size": 4,
|
| 48 |
+
"path_norm": true,
|
| 49 |
+
"pretrained_window_sizes": [
|
| 50 |
+
12,
|
| 51 |
+
12,
|
| 52 |
+
12,
|
| 53 |
+
6
|
| 54 |
+
],
|
| 55 |
+
"problem_type": "single_label_classification",
|
| 56 |
+
"qkv_bias": true,
|
| 57 |
+
"stage_names": [
|
| 58 |
+
"stem",
|
| 59 |
+
"stage1",
|
| 60 |
+
"stage2",
|
| 61 |
+
"stage3",
|
| 62 |
+
"stage4"
|
| 63 |
+
],
|
| 64 |
+
"transformers_version": "4.57.6",
|
| 65 |
+
"use_absolute_embeddings": false,
|
| 66 |
+
"window_size": 16
|
| 67 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea8282e33616dd8c5aa9f5867ae40e89e0fe0aa87ae98770e378aa3b88642bdd
|
| 3 |
+
size 780886280
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.485,
|
| 8 |
+
0.456,
|
| 9 |
+
0.406
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "ViTImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.229,
|
| 14 |
+
0.224,
|
| 15 |
+
0.225
|
| 16 |
+
],
|
| 17 |
+
"resample": 3,
|
| 18 |
+
"rescale_factor": 0.00392156862745098,
|
| 19 |
+
"size": {
|
| 20 |
+
"height": 256,
|
| 21 |
+
"width": 256
|
| 22 |
+
}
|
| 23 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:11bd20e069855e04cc796ab077650d16f81623411401c5651d4254bc705bc8e7
|
| 3 |
+
size 5777
|